WPC'  
      2       B   /   R   <       K   K  #| D   w Roman CG Times CG Times Italic CG Times Bold Univers XPCG Times CG Times Italic CG Times Bold Univers Univers Bold g99ttt[~rEP~rr~~E9E`gE[g[g[Egg99g9ggggEP9gggg[c)coE3EE"EEE   C EEEEEE dEg9 YYYYYY~Y~Y~Y~YC8C8C8C8ddddddddddYggdddsdd[[[[[[g~[~[~[~[ggggggggE8EgEEEgoPEg~9~E~E~8~9g vggggEEErPkNrPrP~E~E~Egggg  gg~[YNH d3 ]gE[gggg g9@ gFdddgF %C [[gEEggzzddd wrr     E[      d     FCG Times CG Times Italic CG Times Bold Univers Univers Bold Univers Italic  \XP 2   .              l      USUK  3'                                          3'Standard                                  6&                                          6& Standard /   rJet 4   A* D                                            #  Xw     P 7XP# Times New Roman CG Times  "  Sh5^!%377b\%%%7b%%%%7777777777%%nbn7bOIOOICVV+5VIhOVCVO=IOOnOOI%%%77%7=1=1'7=%=\=7==1+%=7O771777n777q777   % 777777 77= O7O7O7O7O7rOO1I1I1I1I1++++O=V7V7V7V7O=O=O=O=O7O7O=V7V7O7O7C=O7O7O7O1O7O7O7O=I7I7I7I7V7V7V7V7V7V7V=V=77=77+=d?77V=II7I7I+IO= XO=O=V7V7|QO1O7O7=+=+=7=7I7I7I7O=O=O=O=  O=nOO7I1I1%     77   %      %%WnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxNOOO7 77%7//77k777bb7'7777b' %% ??7bnn77CCTn7n7O7 E\\==b     %?      7  h   '"nnnnOOnnn7Onn?OObbnnn             7bbnnn b7%%nnnnnOCnbOOOOOOOOOO %n7O7=7bnOnI7%7O17?OVKnnOOOOOOOOOOnnnnn5nOOOOnnnnnnnnnnnnn nnnnnnOOnnnnnn`ICVnOn   nn    /A                                               OOOO                       nn          n        nn         nOOO7OOnOOnnn   OO   O7   OOOO7   OOOOO   OOO   OOO   OOO   OOOOOOOOOOOOOO     O   OOOOOO    OOOOOOOO%   OOO%   OOO%   OOO%   OOOO   O                                                   O   OOOO   OOOOCI=  C=I1I+I1V1O1+V1I=h=O1I7V1OCC7C7 7I1O1ICO=VCCCOIV+7VOOC1C C CCQ%%777 7                      O                         V                   777                OOO         C                  %  C71%x}c3wc}}yMMc>Bwgwgcccccc@XKKgg;;c\\\{IZu__FXOOMKV_aaa}rgweeencwrTpOgOaaaa0Oja>rraaa99!aaaaag0aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaappppllllllllllVVaaaaaajjaaTTTTrraa____aaXXrXTTTT3(DTaa                                                                                                                                                                    d                        Y           Y                                                                                                  C          C          C          C                                                                                                                                              z ~ o     z o  Y ~ N  Y  d  Y C 8  Y  o  o  Y  d  Y  z s d z d   d ~ Y  Y  z  o  z z z  ~  C d    z Y z   z   z z  C C d d d   d                                                                                                                                       d d d                                                      z                                     C     s  d Y C  x    \             p x           t      l l                                           X    p      h h <       X                                                                                  \ H |     2   X       `   Z   h  	       7 i C 3 ,   'Xi \	    P 6  XP  8 w C ; ,  Xw     P E37XP V " G ( $ ,   hG     P E37hP> X   F ( $ ,  `hF &_  x 7hX#| Y   HP LaserJet 4                        HPLAS4.PRS  4Y     P 7 \P 2         l    '   i  v      USUK  3'                                          3'Standard                                  6&                                          6& Standard        HPLAS4.PRS  4Y                                                "  Sh5^,2EKKu222K2222KKKKKKKKKK22Cl^go^Ylo2:j^olVldQ^ollla222KK2CKCKC2KK**K*uKKKK8:*KKlKKCCKCK'KKKKK   2 KKKKKK KKK* lClClClClCdgC^C^C^C^C2*2*2*2*oKlKlKlKlKoKoKoKoKlKlCoKlKlKlKoKVKlKlKlKgCgKgKgKoK^K^K^K^KlKlKlKlKlKlKoKoKK*KKKK2KnSKKjK^*^K^K^*^*oK YoKoKlKlKld8dKdKQ:Q:QKQK^K^K^KoKoKoKoK  oKllKaCaC2     KK   2      22WxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxNlllK KK2C==KKKKKK5KKKK5 22 NNK''KK\\rKKlK Y}}SS     2N      K     5"llKlNll             K充+ K22l\llllllllll 2KlKSKldK2KlCKVluallllllllllEllll ll}d\ul       @Y                                               llll                                                  lllKllll   ll   lC   llllC   lllll   lll   lll   lll   llllllllllllll     l   llllll    llllllll2   lll2   lll2   lll2   llll   l                                                   l   llll   llll\^S  \SdC^:aCoKlC2*jCdSSoCdKlCl\VK\K K^ClCd\lSu\\\l^o2Klll\C\ \ \\n22KKK K                      l                         o                   KKK                lll         \                  2  VKC2xɇEإiiTZWxffQQ~~~c{`xllifurllBlTNN-BuurrrrxxxrrrrE6]rCG Times CG Times Italic CG Times Bold a8Document g        Document Style  Style                                       X X`	`	  `	

 2 #  p   8  k     k        ~  a4Document g        Document Style  Style                                      .  a6Document g        Document Style  Style                                    G  X  

a5Document g        Document Style  Style                                   }    X (#

a2Document g        Document Style  Style                                  < o  
   ?                    A.        

 2 |  v   U  t        ?  	     a7Document g        Document Style  Style                                   y    X  X`	`	 (#`	

Bibliogrphy          Bibliography                                             :   X 
 (#

a1Right Par         Right-Aligned Paragraph Numbers                        : ` S  @                   I.  
  X (#

a2Right Par         Right-Aligned Paragraph Numbers                        	C  	   @`	                  A.    `	`	 (#`	

 2 u  
        U       
     a3Document g        Document Style  Style                                  
B 
 b 
   ?                     1.        
a3Right Par         Right-Aligned Paragraph Numbers                        L ! 
   `	`	 @P
                  1.  `	`	   (#

a4Right Par         Right-Aligned Paragraph Numbers                        U  j   `	`	  @                  a.    `	 (#

a5Right Par         Right-Aligned Paragraph Numbers                        
_ o    `	`	   @h                  (1)    hh# (#h

 2 "          o      ?!     "  a6Right Par         Right-Aligned Paragraph Numbers                        h     `	`	   hh# @$                  (a)  hh#  ( (#

a7Right Par         Right-Aligned Paragraph Numbers                        p fJ    `	`	   hh# ( @*                  i)  (  h- (#

a8Right Par         Right-Aligned Paragraph Numbers                        y W" 3!   `	`	   hh# ( - @p/                  a)  -  pp2 (#p

a1Document g        Document Style  Style                                  X q q
    
   l   ^)                       I.           ׃

 2 O&  +  #     A$     A%     %  Doc Init             Initialize Document Style                                  
 
               0*0*    I. A. 1. a.(1)(a) i) a)                 I. 1. A. a.(1)(a) i) a)                                     Document g                                           Tech Init             Initialize Technical Style                              .  
k    I. A. 1. a.(1)(a) i) a)                 1 .1 .1 .1 .1 .1 .1 .1                                      Technical                                            a5Technical         Technical Document Style                               ) W D                   (1)  .  a6Technical         Technical Document Style                               )  D                   (a)  .   2 >)     &     /'     '     |(  a2Technical         Technical Document Style                               < 6  
   ?                    A.        

 a3Technical         Technical Document Style                               9 W g 
   2                    1.        
 a4Technical         Technical Document Style                               8 bv {    2                     a.        
 a1Technical         Technical Document Style                               F ! < 
   ?                         I.           

  2 /     p)     )    |*  O  .  a7Technical         Technical Document Style                               ( @ D                   i)  .  a8Technical         Technical Document Style                               (  D                   a)  .  Pleading              Header for numbered pleading paper                     P@  n                         $]        X    X`	hp x (#%'0*,.8135@8:<H?A                                         y    *                    d       d d                                                                         y y    *                    d       d d                                                                         y 

HH 1

HH 2

HH 3

HH 4

HH 5

HH 6

HH 7

HH 8

HH 9

H 10

H 11

H 12

H 13

H 14

H 15

H 16

H 17

H 18

H 19

H 20

H 21

H 22

H 23

H 24

H 25

H 26

H 27

H 28	 + 	 Ӌheader    4
                                                        _      4
<D                    !                    #  x \	    P C    XP#  2 
F      0  l  0  
  5  l  @  page number   lly e                                                       '       #  x \	    P C    XP#  "  Sh5^18MSS888S8888SSSSSSSSSS88Jxir{icx{8Aui{x`xoZi{xxxl888SS8JSJSJ8SS..S.SSSS>A.SSxSSJJSJS+SSSSS   8 SSSSSS SSS. xJxJxJxJxJorJiJiJiJiJ8.8.8.8.{SxSxSxSxS{S{S{S{SxSxJ{SxSxSxS{S`SxSxSxSrJrSrSrS{SiSiSiSiSxSxSxSxSxSxS{S{SS.SSSS8Sz]SSuSi.iSiSi.i.{S c{S{SxSxSxo>oSoSZAZAZSZSiSiSiS{S{S{S{S  {SxxSlJlJ8     SS   8      88WxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxNxxxS SS8JDDSSSSSS;SSSS; 88 VVS++SSffSSxS c]]     8V      S     ;"xxSxWxx             S唔0 S88xfxxxxxxxxxx 8SxS]SxoS8SxJS`xlxxxxxxxxxxMxxxx xxofx       Gc                                               xxxx                                                  xxxSxxxx   xx   xJ   xxxxJ   xxxxx   xxx   xxx   xxx   xxxxxxxxxxxxxx     x   xxxxxx    xxxxxxxx8   xxx8   xxx8   xxx8   xxxx   x                                                   x   xxxx   xxxxfi]  f]oJiAlJ{SxJ8.uJo]]{JoSxJxf`SfS SiJxJofx]fffxi{8SxxxfJf f ffz88SSS S                      x                         {                   SSS                xxx         f                  8  `SJ8xߖMuu]daqqZZnkxxuqxxIx]WW2IςM<g"  ` H2	:	^ 6 = U \ \   = = = \  = = = = \ \ \ \ \ \ \ \ \ \ = =    Q   s ~  s m   = G  s    i  z b s      w = = = \ \ = Q \ Q \ Q = \ \ 3 3 \ 3  \ \ \ \ D G 3 \ \  \ \ Q Q \ Q   \ / \ \   \ \ \       =   \ \ \ \ \ \   \ \ \ 3    Q  Q  Q  Q  Q  z ~ Q s Q s Q s Q s Q = 3 = 3 = 3 = 3  \  \  \  \  \  \  \  \  \  \  Q  \  \  \  \  \ i \  \  \  \ ~ Q ~ \ ~ \ ~ \  \ s \ s \ s \ s \  \  \  \  \  \  \  \  \ \ 3 \ \ \ \ = \  f \ \  \ s 3 s \ s \ s 3 s 3  \   m  \  \  \  \   z D z \ z \ b G b G b \ b \ s \ s \ s \  \  \  \  \      \    \ w Q w Q  =           \ \       =             = = W                                                 x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x N    \   \ \ = Q K K \ \  \ \ \   \ A \ \ \ \  A   = =   _ _ \  0 0   \ \ p p   \  \  \   m   f f            = _             \             A "          \    _                                  \       4     \ = =       p               =  \  \ f \     z \ = \  Q \ i   w                  U                                    z p                     N m                                                                                                                                                                                                                 \                        Q           Q                                                                                                  =          =          =          =                                                                                                                                              p s f     p f z Q s G w Q  \  Q = 3  Q z f  f  Q z \  Q  p i \ p \   \ s Q  Q z p  f  p p p  s  = \    p Q p   p   p p  = = \ \ \   \                                                                                                                                       \ \ \                                                      p                                     =     i  \ Q =  x      T            g n           j  } }   c c      y     u     }                           Q    g      _ _ 7       Q                                                                                  T B r     "  Sh5^18PSS888S8888SSSSSSSSSS88Sffoxf`xx8Jo]oxfxfS]xff]]888SS8SSJSJ.SS..J.xSSSSAA.SJoJJAJSJS+SSSSS   8 SSSSSS SSS. fSfSfSfSfSooJfJfJfJfJ8.8.8.8.oSxSxSxSxSxSxSxSxS]JfSxSxSxS]JxSfSfSfSfSoJoSoSoSxSfSfSfSfSxSxSxSxSxSxSxSxSS.SSSS8SSSSoS].]S]S].].oS foSoSxSxSofAfSfSSASASSSS]S]S]SxSxSxSxS  xSo]S]A]A8     SS   8      88WxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxNxxxS SS8SMMSSSSSS;SSSS; 88 SSS..SSffSSxS ZSS     8S      S     ;"xxSxSxx             S唔0 S88xfxxxxxxxxxx 8SfS]SxoS8SxJS`xlxxxxxxxxxxMxxxx xxofx       Gc                                               xxxx                                                  xxxSxxxx   xx   xJ   xxxxJ   xxxxx   xxx   xxx   xxx   xxxxxxxxxxxxxx     x   xxxxxx    xxxxxxxx8   xxx8   xxx8   xxx8   xxxx   x                                                   x   xxxx   xxxfff]  f]oJfA]JxSxJ8.oJo]]oJoSxJxffSfS S]J]Joff]fffffx8Sx]]fJf f ffz88SSS S                      f                         x                   SSS                ]]]         f                  8  `SJ8xߖMuu]daqqZZnkxxuqxxIx]WW2IςM<g 2 f  
  ?F  
  P  l  [  l  a  "  ` H2	:	^ 6 = U \ \   = = = \  = = = = \ \ \ \ \ \ \ \ \ \ = =    \   z   z p   G X  z    p   f z      z = = = \ \ = \ f Q f Q A \ f 3 = f 3  f \ f f Q G = f \  \ \ Q \ \ \   \  \ \   \ \ \       =   \ \ \ \ \ \   \ \ f 3    \  \  \  \  \    Q z Q z Q z Q z Q G 3 G 3 G 3 G 3  f  \  \  \  \  f  f  f  f  \  \  f  \  \  \  \ p f  \  \  \  Q  \  \  \  f z \ z \ z \ z \  \  \  \  \  \  \  f  f \ 3 \ f \ \ G f  i \ \  f z 3 z \ z \ z G z 3  f     f  f  \  \    Q  \  \ f G f G f \ f \ z \ z \ z \  f  f  f  f      f    \ z Q z Q  =           \ \       =             = = W                                                 x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x N    \   \ \ = \ N N \ \  \ \ \   \ A \ \ \ \  A   = =   i i \  0 0   \ \ p p   \  \  \   s   f f            = i             \             A "          \    i                                  \       4     \ = =       p               =  \  \ f \     z \ = \  Q \ i   ~                  X                                    z p                     N m                                                                                                                                                                                                                 \                        \           \                                                                                                  =          =          =          =                                                                                                                                              p z f     p f z Q z G z Q  Q  Q G 3  Q z f  f  Q z \  Q  p p \ p \   \ z Q  Q z p  f  p p p  z  G \    p Q p   p   p p  = = \ \ \   \                                                                                                                                       \ \ \                                                      p                                     =     p  \ Q =  x      T            g n           j  } }   c c      y     u     }                           Q    g      _ _ 7       Q                                                                                  T B r    "  ` H2	:	^ 6 = X \ \   = = = \  = = = = \ \ \ \ \ \ \ \ \ \ = =    \  p p z  p i   = Q z f  z  p  p \ f  p  p f f = = = \ \ = \ \ Q \ Q 3 \ \ 3 3 Q 3  \ \ \ \ G G 3 \ Q z Q Q G Q \ Q   \ / \ \   \ \ \       =   \ \ \ \ \ \   \ \ \ 3   p \ p \ p \ p \ p \  z z Q p Q p Q p Q p Q = 3 = 3 = 3 = 3 z \  \  \  \  \  \  \  \  \ f Q p \  \  \  \ f Q  \ p \ p \ p \ p \ z Q z \ z \ z \  \ p \ p \ p \ p \  \  \  \  \  \  \  \  \ \ 3 \ \ \ \ = \  \ \ \ z \ f 3 f \ f \ f 3 f 3 z \   p z \ z \  \  \  z p G p \ p \ \ G \ G \ \ \ \ f \ f \ f \  \  \  \  \      \  z f \ f G f G  =           \ \       =             = = W                                                 x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x N    \   \ \ = \ U U \ \  \ \ \   \ A \ \ \ \  A   = =   \ \ \  3 3   \ \ p p   \  \  \   b   \ \            = \             \             A "          \    [                                  \       4     \ = =       p               =  \ p \ f \     z \ = \  Q \ i   w                  U                                    z p                     N m                                                                                                                                                                                                                 \                        Q           Q                                                                                                  =          =          =          =                                                                                                                                             p p p f     p f z Q p G f Q  \  Q = 3 z Q z f  f z Q z \  Q  p p \ p \   \ f Q f Q z p p f  p p p p p  = \  f f p Q p   p   p p  = = \ \ \   \                                             p                                                                                          \ \ \                                 f f f                   p                                     =     i  \ Q =  x      T            g n           j  } }   c c      y     u     }                           Q    g      _ _ 7       Q                                                                                  T B r     "  Sh5^!%377bV%%%7b%%%%7777777777%%nbn1bOEKQEAOQ%+MEdQO?OI;EQOhOOG%%%77%17171%777V7777)+77O771171n777q777   % 777777 777 O1O1O1O1O1bIK1E1E1E1E1%%%%Q7O7O7O7O7Q7Q7Q7Q7O7O1Q7O7O7O7Q7?7O7O7O7K1K7K7K7Q7E7E7E7E7O7O7O7O7O7O7Q7Q777777%7P=77M7EE7E7EEQ7 AQ7Q7O7O7bOI)I7I7;+;+;7;7E7E7E7Q7Q7Q7Q7  Q7hOO7G1G1%     77   %      %%WnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxNOOO7 77%1--77a777bb7'7777b' %% 997bnn77CCTn7n7O7 A\\==b     %9      7  h   '"nnnnOOnnn7Onn9OObbnnn             7bbnnn b7%%nnnnnOCnbOOOOOOOOOO %n7O7=7bnOnI7%7O17?OVGnnOOOOOOOOOOnnnnn3nOOOOnnnnnnnnnnnnn nnnnnnOOnnnnnn\ICVnOn   nn    /A                                               OOOO                       nn          n        nn         nOOO7OOnOOnnn   OO   O1   OOOO1   OOOOO   OOO   OOO   OOO   OOOOOOOOOOOOOO     O   OOOOOO    OOOOOOOO%   OOO%   OOO%   OOO%   OOOO   O                                                   O   OOOO   OOOOCE=  C=I1E+G1Q7O1%M1I=d=Q1I7O1OC?7C7 7E1O1ICO=VCCCOEQ%7OOOC1C C CCQ%%777 7                      O                         Q                   777                OOO         C                  %  ?71%x}c3wc}}yMMc>Bwgwgcccccc@XKKgg;;c\\\{IZu__FXOOMKV_aaa}rgweeencwrTpOgOaaaa0Oja>rraaa99!aaaaag0aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaappppllllllllllVVaaaaaajjaaTTTTrraa____aaXXrXTTTT3(DTaa "  Sh5^+1DIIr111I1111IIIIIIIIII11Aj\dm\Wjm19g\mjTjbO\mjjj_111II1AIAIA1II))I)rIIII69)IIjIIAAIAI&IIIII   1 IIIIII III) jAjAjAjAjAbdA\A\A\A\A1)1)1)1)mIjIjIjIjImImImImIjIjAmIjIjIjImITIjIjIjIdAdIdIdImI\I\I\I\IjIjIjIjIjIjImImII)IIII1IkQIIgI\)\I\I\)\)mI WmImIjIjIjb6bIbIO9O9OIOI\I\I\ImImImImI  mIjjI_A_A1     II   1      11WxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxNjjjI II1A<<IIIIII4IIII4 11 LLI&&IIZZoIIjI WzzQQ     1L      I     4"jjIjLjj             I傂* I11jZjjjjjjjjjj 1IjIQIjbI1IjAITjr_jjjjjjjjjjDjjjj jjzbZrj       >W                                               jjjj                                                  jjjIjjjj   jj   jA   jjjjA   jjjjj   jjj   jjj   jjj   jjjjjjjjjjjjjj     j   jjjjjj    jjjjjjjj1   jjj1   jjj1   jjj1   jjjj   j                                                   j   jjjj   jjjjZ\Q  ZQbA\9_AmIjA1)gAbQQmAbIjAjZTIZI I\AjAbZjQrZZZj\m1IjjjZAZ Z ZZl11III I                      j                         m                   III                jjj         Z                  1  TIA1xńCӡggRXUuddOO{{{ax~~^ujjgdr~ojjAjRLL,Arroooo~~~~uuuooooC5[o 2   l  f  	l  !l  
l  q  
  v   "  Sh5^18MSS888S8888SSSSSSSSSS88SxoxxofAPoxfx]oxxxxo888SS8S]J]J;S].8].]S]]JA8]SxSSJSSSSSSSSS   8 SSSSSS SS]. xSxSxSxSxSxxJoJoJoJoJA.A.A.A.x]SSSSx]x]x]x]xSxSx]SSxSxSf]xSxSxSxJxSxSxSx]oSoSoSoSSSSSSS]]S.S]SSA]_SS]o.oSoSoAo.x] x]x]SS{xJxSxS]A]A]S]SoSoSoSx]x]x]x]  x]xxSoJoJ8     SS   8      88WxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxNxxxS SS8SGGSSSSSS;SSSS; 88 ``S++SSffSSxS i]]     8`      S     ;"xxSx`xx             S唔0 S88xfxxxxxxxxxx 8SxS]SxoS8SxJS`xrxxxxxxxxxxPxxxx xxofx       Gc                                               xxxx                                                  xxxSxxxx   xx   xS   xxxxS   xxxxx   xxx   xxx   xxx   xxxxxxxxxxxxxx     x   xxxxxx    xxxxxxxx8   xxx8   xxx8   xxx8   xxxx   x                                                   x   xxxx   xxxxfo]  f]oJoAoJJxJA.Jo]]xJoSJxffSfS SoJxJofx]fffxoASxxfJf f ffz88SSS S                      x                                            SSS                xxx         f                  8  fSJ8xߖMuu]daqqZZnkxxuqxxIx]WW2IςM<g "  Sh5^,2EKK}222K2222KKKKKKKKKK22Kldlld\uu:Hudlu\ulSdlllld222KK2KSCSC5KS*2S*}SKSSC:2SKlKKCKKKKKKKKK   2 KKKKKK KKS* lKlKlKlKlKllCdCdCdCdC:*:*:*:*lSuKuKuKuKlSlSlSlSlKlKlSuKuKlKlK\SlKlKlKlClKlKlKlSdKdKdKdKuKuKuKuKuKuKuSuSK*KSKK:SVKKuSd*dKdKd:d*lS wlSlSuKuKolClKlKS:S:SKSKdKdKdKlSlSlSlS  lSllKdCdC2     KK   2      22WxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxNlllK KK2K@@KKKKKK5KKKK5 22 VVK''KK\\rKKlK ^}}SS     2V      K     5"llKlVll             K充+ K22l\llllllllll 2KlKSKldK2KlCKVlugllllllllllHllll lld\ul       @Y                                               llll                                                  lllKllll   ll   lK   llllK   lllll   lll   lll   lll   llllllllllllll     l   llllll    llllllll2   lll2   lll2   lll2   llll   l                                                   l   llll   llll\dS  \SdCd:dCuClC:*uCdSSlCdKuCl\\K\K KdClCd\lSu\\\ldu:Kull\C\ \ \\n22KKK K                      l                         u                   KKK                lll         \                  2  \KC2xɇEإiiTZWxffQQ~~~c{`xllifurllBlTNN-BuurrrrxxxrrrrE6]r "  Sh5^,2HKKu222K2222KKKKKKKKKK22K\\dl\Vll2CdS}dl\l\KSl\}\SS222KK2KKCKC*KK**C*lKKKK::*KCdCC:CKCK'KKKKK   2 KKKKKK KKK* \K\K\K\K\KddC\C\C\C\C2*2*2*2*dKlKlKlKlKlKlKlKlKSC\KlKlKlKSClK\K\K\K\KdCdKdKdKlK\K\K\K\KlKlKlKlKlKlKlKlKK*KKKK2KzKKKdKS*SKSKS*S*dK \dKdKlKlKd\:\K\KK:K:KKKKSKSKSKlKlKlKlK  lK}dSKS:S:2     KK   2      22WxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxNlllK KK2KEEKKKKKK5KKKK5 22 KKK**KK\\rKKlK QrrKK     2K      K     5"llKlKll             K充+ K22l\llllllllll 2K\KSKldK2KlCKVluallllllllllEllll ll}d\ul       @Y                                               llll                                                  lllKllll   ll   lC   llllC   lllll   lll   lll   lll   llllllllllllll     l   llllll    llllllll2   lll2   lll2   lll2   llll   l                                                   l   llll   lll\\\S  \SdC\:SClKlC2*dCdS}SdCdKlCl\\K\K KSCSCd\\Su\\\\\l2KlSS\C\ \ \\n22KKK K                      \                         l                   KKK                SSS         \                  2  VKC2xɇEإiiTZWxffQQ~~~c{`xllifurllBlTNN-BuurrrrxxxrrrrE6]r"  ` H2	:	^ ; C ] d d   C C C d  C C C C d d d d d d d d d d C C    Y   ~   ~ w   C N  ~    s   k ~       C C C d d C Y d Y d Y C d d 8 8 d 8  d d d d J N 8 d d  d d Y Y d Y   d 3 d d  " d d d       C   d d d d d d   d d d 8    Y  Y  Y  Y  Y    Y ~ Y ~ Y ~ Y ~ Y C 8 C 8 C 8 C 8  d  d  d  d  d  d  d  d  d  d  Y  d  d  d  d  d s d  d  d  d  Y  d  d  d  d ~ d ~ d ~ d ~ d  d  d  d  d  d  d  d  d d 8 d d d d C d  o d d  d ~ 8 ~ d ~ d ~ 8 ~ 8  d   v  d  d  d  d    J  d  d k N k N k d k d ~ d ~ d ~ d  d  d  d  d      d    d  Y  Y  C           d d       C             C C W                                                 x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x N    d   d d C Y Q Q d d  d d d   d F d d d d  F   C C   h h d  4 4   d d z z   d  d  d   w   o o            C h             d             F "          d    h                                  d       9     d C C       z               C  d  d o d      d C d  Y d s                     ]                                     z                     U w                                                                                                                                                                                                                 d                        Y           Y                                                                                                  C          C          C          C                                                                                                                                              z ~ o     z o  Y ~ N  Y  d  Y C 8  Y  o  o  Y  d  Y  z s d z d   d ~ Y  Y  z  o  z z z  ~  C d    z Y z   z   z z  C C d d d   d                                                                                                                                       d d d                                                      z                                     C     s  d Y C  x    \             p x           t      l l                                           X    p      h h <       X                                                                                  \ H |     2 c  l    

  M  `       J    "  Sh5^/5JPP|555P5555PPPPPPPPPP55Gtenwe_tw5>qewt\tkVewttth555PP5GPGPG5PP,,P,|PPPP;>,PPtPPGGPGP)PPPPP   5 PPPPPP PPP, tGtGtGtGtGknGeGeGeGeG5,5,5,5,wPtPtPtPtPwPwPwPwPtPtGwPtPtPtPwP\PtPtPtPnGnPnPnPwPePePePePtPtPtPtPtPtPwPwPP,PPPP5PuYPPqPe,ePePe,e,wP _wPwPtPtPtk;kPkPV>V>VPVPePePePwPwPwPwP  wPttPhGhG5     PP   5      55WxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxNtttP PP5GAAPPPPPP8PPPP8 55 SSP))PPbbyPPtP _YY     5S      P     8"ttPtStt             P厎. P55tbtttttttttt 5PtPYPtkP5PtGP\t|httttttttttJtttt ttkb|t       D_                                               tttt                                                  tttPtttt   tt   tG   ttttG   ttttt   ttt   ttt   ttt   tttttttttttttt     t   tttttt    tttttttt5   ttt5   ttt5   ttt5   tttt   t                                                   t   tttt   ttttbeY  bYkGe>hGwPtG5,qGkYYwGkPtGtb\PbP PeGtGkbtY|bbbtew5PtttbGb b bbu55PPP P                      t                         w                   PPP                ttt         b                  5  \PG5x֐JppZ`]mmVVjfsspm}zssFsZSS0F}}zzzzzzzzJ:cz"  ` H2	:	^ ; C ] d d   C C C d  C C C C d d d d d d d d d d C C    d       z   N `      z   o        C C C d d C d o Y o Y F d o 8 C o 8  o d o o Y N C o d  d d Y d d d   d  d d  " d d d       C   d d d d d d   d d o 8    d  d  d  d  d    Y  Y  Y  Y  Y N 8 N 8 N 8 N 8  o  d  d  d  d  o  o  o  o  d  d  o  d  d  d  d z o  d  d  d  Y  d  d  d  o  d  d  d  d  d  d  d  d  d  d  o  o d 8 d o d d N o  r d d  o  8  d  d  N  8  o     o  o  d  d    Y  d  d o N o N o d o d  d  d  d  o  o  o  o      o    d  Y  Y  C           d d       C             C C W                                                 x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x N    d   d d C d U U d d  d d d   d F d d d d  F   C C   s s d  4 4   d d z z   d  d  d   ~   o o            C s             d             F "          d    s                                  d       9     d C C       z               C  d  d o d      d C d  Y d s                     `                                     z                     U w                                                                                                                                                                                                                 d                        d           d                                                                                                  C          C          C          C                                                                                                                                              z  o     z o  Y  N  Y  Y  Y N 8  Y  o  o  Y  d  Y  z z d z d   d  Y  Y  z  o  z z z    N d    z Y z   z   z z  C C d d d   d                                                                                                                                       d d d                                                      z                                     C     z  d Y C  x    \             p x           t      l l                                           X    p      h h <       X                                                                                  \ H |     l * Y 2 , ,   !1Y     P E37P
	  3 m = 6 ,   b&m     P E37&P x / c 8 1 ,   Apc     P E37P> z - b 8 1 , 	 ib &_  x 7X   2 p = 6 ,  L&p _    p i7&>  1 k = 6 , 	 &k &_  x 7&X O  A % ! ,   JA     P E37JP j ) W 1 + ,   PW     P E37P  y . f 8 1 ,  f _    p i7  m ) \ 2 , ,  \ _    p i7> n ( X 2 , , 	 (X &_  x 7X  8 w C ; ,   Xw     P E37XP
 s - _ 5 / ,   _     P E37P   7 z C ; ,  c!Xz _    p i7X  O  C % ! ,  @pJC _    p i7J  j ) Y 1 + ,  `Y _    p i7 2   l    
    b    '       "  Sh5^+1DIIz111I1111IIIIIIIIII11IjbjjbZrr9GrbjrZrjQbjjjjb111II1IQAQA4IQ)1Q)zQIQQA91QIjIIAIIIIIIIII   1 IIIIII IIQ) jIjIjIjIjIjjAbAbAbAbA9)9)9)9)jQrIrIrIrIjQjQjQjQjIjIjQrIrIjIjIZQjIjIjIjAjIjIjIjQbIbIbIbIrIrIrIrIrIrIrQrQI)IQII9QTIIrQb)bIbIb9b)jQ ujQjQrIrImjAjIjIQ9Q9QIQIbIbIbIjQjQjQjQ  jQjjIbAbA1     II   1      11WxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxNjjjI II1I>>IIIIII4IIII4 11 TTI&&IIZZoIIjI \zzQQ     1T      I     4"jjIjTjj             I傂* I11jZjjjjjjjjjj 1IjIQIjbI1IjAITjrdjjjjjjjjjjGjjjj jjbZrj       >W                                               jjjj                                                  jjjIjjjj   jj   jI   jjjjI   jjjjj   jjj   jjj   jjj   jjjjjjjjjjjjjj     j   jjjjjj    jjjjjjjj1   jjj1   jjj1   jjj1   jjjj   j                                                   j   jjjj   jjjjZbQ  ZQbAb9bArAjA9)rAbQQjAbIrAjZZIZI IbAjAbZjQrZZZjbr9IrjjZAZ Z ZZl11III I                      j                         r                   III                jjj         Z                  1  ZIA1xńCӡggRXUuddOO{{{ax~~^ujjgdr~ojjAjRLL,Arroooo~~~~uuuooooC5[o"  ` H2	:	^ ; C ` d d   C C C d  C C C C d d d d d d d d d d C C    d  z z   z s   C Y  o    z  z d o  z  z o o C C C d d C d d Y d Y 8 d d 8 8 Y 8  d d d d N N 8 d Y  Y Y N Y d Y   d 3 d d  " d d d       C   d d d d d d   d d d 8   z d z d z d z d z d    Y z Y z Y z Y z Y C 8 C 8 C 8 C 8  d  d  d  d  d  d  d  d  d o Y z d  d  d  d o Y  d z d z d z d z d  Y  d  d  d  d z d z d z d z d  d  d  d  d  d  d  d  d d 8 d d d d C d  d d d  d o 8 o d o d o 8 o 8  d   z  d  d  d  d   z N z d z d d N d N d d d d o d o d o d  d  d  d  d      d   o d o N o N  C           d d       C             C C W                                                 x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x N    d   d d C d ] ] d d  d d d   d F d d d d  F   C C   d d d  8 8   d d z z   d  d  d   k   d d            C d             d             F "          d    d                                  d       9     d C C       z               C  d z d o d      d C d  Y d s                     ]                                     z                     U w                                                                                                                                                                                                                 d                        Y           Y                                                                                                  C          C          C          C                                                                                                                                             z z z o     z o  Y z N o Y  d  Y C 8  Y  o  o  Y  d  Y  z z d z d   d o Y o Y  z z o  z z z z z  C d  o o z Y z   z   z z  C C d d d   d                                             z                                                                                          d d d                                 o o o                   z                                     C     s  d Y C  x    \             p x           t      l l                                           X    p      h h <       X                                                                                  \ H |     l * Y 2 , ,   !1Y     P E37P
	  3 m = 6 ,   b&m     P E37&P x / c 8 1 ,   Apc     P E37P> z - b 8 1 , 	 ib &_  x 7X   2 p = 6 ,  L&p _    p i7&>  1 k = 6 , 	 &k &_  x 7&X O  A % ! ,   JA     P E37JP j ) W 1 + ,   PW     P E37P  y . f 8 1 ,  f _    p i7  m ) \ 2 , ,  \ _    p i7> n ( X 2 , , 	 (X &_  x 7X  8 w C ; ,   Xw     P E37XP
 s - _ 5 / ,   _     P E37P   7 z C ; ,  c!Xz _    p i7X  O  C % ! ,  @pJC _    p i7J  j ) Y 1 + ,  `Y _    p i7>  6 u C ; , 	 cXu &_  x 7XX6} * i 8 8 , ' i x    P 7Pt} * j 8 8 , / CVj     p s7 2 #  l  K  g     l          "  Sh5^88Sii888i8888iiiiiiiiii88V{iux`]{x.]oZxcllixxxrf888SS8Z`Z`Z>``((Z(````;S>`]]]PJSJiiiiii   8 iiiiii Sif( {Z{Z{Z{Z{ZuZ`Z`Z`Z`Z.(.(.(.(x`````x`x`x`x`r]{Zxi``r]x`c`{Z{i{iuZuiuiuixiiiiiiiii{`{`{i{i{i{ixixii(iiii.fPiioiZ(iiiiZ=Z(x` rxixiiil;lililSlSliliiiiiiix`xixixi  x`rifPfP8     SS   8      88WxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxNxxxS ii8V]]iiiDDiAASSiA 88 SSi>>iiffSSxS rff     8S      S     A"xxSxXxx             S姧0	 S88xcxxxxxxxxxx 8S{i]ix{S8ixSi`xlxxxxxxxxxxZxxxx xxofx       Gc                                               xxxx                                                  xxxSxxxx   xx   xJ   xxxxJ   xxxxx   xxx   xxx   xxx   xxxxxxxxxxxxxx     x   xxxxxx    xxxxxxxx8   xxx8   xxx8   xxx8   xxxx   x                                                   x   xxxx   xxx{]iZ  ]S{Z`MfGx`Z.(oS{V]x]iG`x`cZcc JiMrZuxPr{{`x8irr{Z]  rrz88iii i                      {                         x                   iii                rrr         {                  8  `SJ8xߖMuu]daqqZZnkxxuqxxIx]WW2IςM<g "  Sh5^88Vii888i8888iiiiiiiiii88V{iux`]{x.]oZxcllixxxrf888SS8Z`Z`Z>``((Z(````;S>`]]]PSSSiiiiii   8 iiiiii Sif( {Z{Z{Z{Z{ZuZ`Z`Z`Z`Z.(.(.(.(x`````x`x`x`x`r]{Zxi``r]x`c`{Z{i{iuZuiuiuixiiiiiiiii{`{`{i{i{i{ixixii(iiii.fPiioiZ(iiiiZAZ.x` xxixiiil;lililSlSliliiiiiiix`xixixi  x`rifPfP8     SS   8      88WddddddddddddddddddddddddddddddddddddddddNxxxS ii8V]]iiiDDiAASSiA 88 SSi>>iiffSSxS rff     8S      S     A"xxSx]xx             S姧0	 S88xfxxxxxxxxxx 8S{ici{P8ix]i`xrxxxxxxxxxxSxxxx xxofx       Gc                                               xxxx                                                  xxxSxxxx   xx   xS   xxxxS   xxxxx   xxx   xxx   xxx   xxxxxxxxxxxxxx     x   xxxxxx    xxxxxxxx8   xxx8   xxx8   xxx8   xxxx   x                                                   x   xxxx   xxx{`ic  ]S{``SfMx`f.+oS{Scx]xP`fc`fl MiMrcxx]x{`x8irr`l  rxz88iii i                      {                         x                   iii                rrr                           8  fSJ8xߖMuu]daqqZZnkxxuqxxIx]WW2IςM<g 2     l  U   d        %             "  Sh5^88Sii888i8888iiiiiiiiii88V{iux`]{x.]oZxcllixxxrf888SS8Z`Z`Z>``((Z(````;S>`]]]PJSJiiiiii   8 iiiiii Sif( {Z{Z{Z{Z{ZuZ`Z`Z`Z`Z.(.(.(.(x`````x`x`x`x`r]{Zxi``r]x`c`{Z{i{iuZuiuiuixiiiiiiiii{`{`{i{i{i{ixixii(iiii.fPiioiZ(iiiiZ4Z(x` {xixiiil;lililSlSliliiiiiiix`xixixi  x`rifPfP8     SS   8      88WddddddddddddddddddddddddddddddddddddddddNxxxS ii8V]]iiiDDiAASSiA 88 SSi>>iiffSSxS rff     8S      S     A"xxSx[xx             S姧0 S88xcxxxxxxxxxx 8S{i]ix{S8ixSi`xlxxxxxxxxxxZxxxx xxofx       Gc                                               xxxx                                                  xxxSxxxx   xx   xJ   xxxxJ   xxxxx   xxx   xxx   xxx   xxxxxxxxxxxxxx     x   xxxxxx    xxxxxxxx8   xxx8   xxx8   xxx8   xxxx   x                                                   x   xxxx   xxx{]iZ  ]S{Z`MfGx`Z.(oS{V]x]iG`x`cZcc JiMrZuxPr{{`x8irr{Z]  rrz88iii i                      {                         x                   iii                rrr         {                  8  `SJ8xߖMuu]daqqZZnkxxuqxxIx]WW2IςM<g l * Y 2 , ,   !1Y     P E37P
	  3 m = 6 ,   b&m     P E37&P x / c 8 1 ,   Apc     P E37P> z - b 8 1 , 	 ib &_  x 7X   2 p = 6 ,  L&p _    p i7&>  1 k = 6 , 	 &k &_  x 7&X O  A % ! ,   JA     P E37JP j ) W 1 + ,   PW     P E37P  y . f 8 1 ,  f _    p i7  m ) \ 2 , ,  \ _    p i7> n ( X 2 , , 	 (X &_  x 7X  8 w C ; ,   Xw     P E37XP
 s - _ 5 / ,   _     P E37P   7 z C ; ,  c!Xz _    p i7X  O  C % ! ,  @pJC _    p i7J  j ) Y 1 + ,  `Y _    p i7>  6 u C ; , 	 cXu &_  x 7XX6} * i 8 8 , ' i x    P 7Pt} * j 8 8 , / CVj     p s7} * i 8 8 , < i   x U7X  f                                                                                                                  S J 8                                           
     
 #  &m     P 7b&P# 
     
 
















a3 Section 2



a Practical Lessons Using Marine and Coastal Image Data


















   !         p-p-p-  
             p-p-p-  
    
  z         q     P  z N    # c     P 7ApP# Marine and Coastal Image Data Module 4:z  p         r     ޕ  z P    # c     P E37ApP# `"l Practical Lessons: 1p # &m     P 7b&P# 
a4 Practical Lessons Using Marine Image Data


   S ` a- 1: IMAGING INSITU BIOOPTICAL AND PHYSICAL PARAMETERSă



   T   aB Aim of Lessonă

To use colour imaging to represent time series from automated sensors moored at several different
depths.



   T  aC Objectivesă

On completion of this lesson, you should be able to:

1) display velocity, chlorophyll fluorescence, and temperature time series

2) observe prominent scales of processes including: seasonal, synoptic, diurnal, and inertial

3) identify physical processes that force biological processes.



   S  ~a= Background Informationă
   S  Equipment

   T R This lesson is an introduction to data imaging of insitu measurements from moored sensors.  Data
were collected during the Biowatt II experiment from March to November, 1987 in the Sargasso Sea
(Figure 1).  Biowatt II was designed to measure upper ocean bio-optical and physical processes using
a unique instrument.  The instrument (named the multi-variable moored system or MVMS), described
   T  in detail by Dickey et al. (1991), consisted of a batterypowered vector measuring current meter (to
record horizontal current speeds and directions, and also temperatures) connected to a fluorometer
(calibrated to chlorophyll), a transmissometer (for estimates of particle abundance), a dissolved
oxygen sensor, and a PAR (photosynthetically available radiation) sensor.  During Biowatt II, MVMS
instruments were moored at eight depths ranging from 10m to 160m, each parameter being recorded
every 4 minutes for approximately 9 months.  

   S " The Sargasso Sea

The Sargasso Sea is located north of the subtropical convergence zone within and to the west of the
North Atlantic gyre (the gyre is bounded by the Gulf Stream to the west and north, the Canary
current to the east and the Equatorial counter current to the south).  The Sargasso Sea has been the
focus of many oceanographic programs (Figure 1).  The mooring site (70$W and 34$N) was
approximately 350 km to the south of the mean axis of the Gulf Stream.  This area is called the
recirculation region, named for the many cyclonic and anticyclonic rings or eddies which propagate
towards the southwest from the northern edge of the Gulf Stream .   ^)        p->*>*BB  ԌIn general we can characterize physical and bio-optical data from moored instruments, in terms of
   T   time scales (Dickey et al. 1992).  For the images in this lesson (taken during the fall of 1987), time
scales include seasonal (a result of the annual heating and cooling), synoptic (a result of storms and
mesoscale eddies), diurnal (resulting from daily radiation fluxes), and inertial (resulting from the
Coriolis effect).  Fall, 1987 in the Sargasso Sea, is characterized by kinetic energy (velocity)
increases, reduced solar radiation, a deepening mixed layer (to below 70m), upward movement of
nutrientrich water into the euphotic zone (which remains at 110m), and a displacement of high
chlorophyll values towards the surface (a fall bloom).  

   S  The images and data processing

Three colour images showing: 1) N-S current speeds, 2) temperature, and 3) chlorophyll fluorescence,
all from a 56day period in the fall of 1987, are used in this lesson.  The images were created by first
filtering the raw time series for each variable with a Gaussian filter (3 hour moving window) and then
averaging the filtered data to 2.67 hour bins to produce 512 horizontal pixels (9 pixels/day).  Vertical
pixels were interpolated every 0.6m for each time point using a cubic spline function fit to all eight
depths (a total of 251 pixels).  The magnitude of each scalar variable was scaled to an integer ranging
from 255 (black-red) to 0 (yellow-green).  The images all have the same format, namely 251 vertical
depth bins (6m/10 pixels) by 512 horizontal time bins (1day/9 pixels).  Table 1 lists the file name of
the three images and describes their contents.  Each image has three files that must be loaded: the data
file, the look-up table (LUT) and the palette, which is common to each image. 



   T B *a; Table 1. Files and contentsă

^  
    d d x                           
 ! d d x                         ^   
 
 
 
              6 Data Files   s       

. Parameter   s       ~hG Range   s       DH\ Colour   s       Ys LUT  
 
8 	 	               FALVV56D          north currents
south currents         hE +80 cm/s
hE 80 cm/s         yH] green
H^ red         q VV.STR 8 	 	8 	 	s               FALTP56D          temperature
(isotherms)         hG 28$C
hG 18$C         H^ red
H] blue         q TP.STR 8 	 	h		               FALFL56D   s       chlorophyll
fluorescence   K         T k hE 2.0 mgmé3ă
   T C 1hF 0 mgmé3   K      yH] green
H] black   K      q FL.STR h		    


   T   The palette file provided is named BW.PAL.

   "        p->*>*BB  
   T   aA Lesson Outlineă

   S  A.  Chlorophyll image 
1.  Type UNESCO13 to enter the imaging software.  

   T  2.  Load the file FALFL56D using the load image option.  Next, choose the load LUT option, type
   T  FL, and press return (this loads the look-up table file, FL.STR).  Finally, select the load palette
   T  option and load BW, which is the palette file (bw.pal) used for all three images.  (Note bw.pal need
only be loaded once since it is used for all three images, however, the LUT must be loaded each time
a new data file is loaded - see Table 1 for the correct LUT).  

   T $
 3.  View the image using the display option.  The colour range has been enhanced so the high
   T 
 chlorophyll values (2mgmé3) are green (integer value of 60) and low are red-black (integer value of
5).  The next task is to determine the depth of the chlorophyll maximum.  The easiest way to
   T  determine this is to use the transect option described below.

   T ^ 4.  Enter the transect option.  Move the topmost cursor to the top left corner of the image and the
other to the bottom left corner.

5.  Move both cursors 100 pixels to the right, which is equivalent to 11 days, that is 100*(1days/9
pixels).  If you have done everything right, the cursors should be aligned on the same time pixel. 
   T  If so, use the space bar to draw a vertical transect and press the enter key to display the graph.  The
resulting graph is a chlorophyll profile with the 10m pixel to the left and the 160m pixel to the far
right.  (If the image is reversed, it is because the cursors are reversed).  Note the subsurface
maximum (large integer values) and lower chlorophyll values in the surface layers.  To escape back
to the colour image, and at the same time save the graph, press either the space bar or return key. 
By toggling between the graphic screen and the colour image, several graphs can be overlaid on the
original image.  Try plotting other profiles to determine if the profile varies in time.  Unfortunately,
the depth of the maximum cannot be estimated accurately from this graph so another method has to
be used.

   T  6.  Without leaving the transect plot, press the <ESC> key to exit directly to the main menu. 
   T  Immediately re-enter the transect option.  This is done to clear the transect screen of previous graphs
while keeping the cursors aligned.  Both cursors should retain the positions they had in the last graph
(ie. the top and bottom of the image and 100 pixels to the right).  Now move the top cursor down
   T j 80 pixels, draw the transect and graph it.  Note that this graph is drawn from the top cursor to the
bottom cursor and that the top cursor is at a depth of 58m or 80*(6m/10 pixels)+10m.  This is the
minimum depth of the deep chlorophyll maximum (DCM).  Keep moving the top cursor downward
and re-plotting profiles until you have determined the depth of the maximum.  

Questions:
A1.  At what depth is the chlorophyll maximum?  
A2.  To what depth does the DCM extend (ie. the maximum depth)?  
A3.  Does the DCM change over time (ie. is the distribution in a steady state)?

   T & 7.  Escape transect and re-enter it.  Now position both cursors at the top left of the image.  Move
both down 20 pixels and one to the right 100 pixels. Draw the horizontal transect and display the
resulting time series.  The periodicity in chlorophyll is difficult to see in this surface mixed layer
since the variability is small (as is the mean chlorophyll value).     d)        p->*>*BB  Ԍ8. Move both cursors down an additional 80 pixels (ie. 100 pixels from the top), draw the transect
and display the graph.  

Questions:
A4.  What is the dominant period of variability in this graph, that is how many peaks/day are
evident?  
A5.  What would explain this periodicity?  Note that there is considerable variability on scales less
than 1 day.

   T  9.  Escape and re-enter transect.  Move both cursors to the top left of the image.  Move one cursor
down 100 pixels; move the other cursor down 180 pixels and 500 pixels to the right; draw the
transect.  Note the slope of this line approximately follows the deepening of the chlorophyll
maximum.  

Question:
A6. What is the rate at which the DCM deepens (in meters/day)?

   T Y B.  Temperature image 
   T 2 1.  Load the file FALTP56D, the LUT TP, and display the image.  The colder water is blue and has
an integer value of 150 and the warmer is red with a value of 50.  The temperature change is 1$C/5
integers (or colour intensities).  Note the number and thicknesses of the isotherms change over time.

   T  2.  Enter the transect option.  Move the topmost cursor to the top left of the image and the other to
the top left and to the right 500 pixels.  Draw the graph at this depth (10m).  The time scale is
exactly the same as that of the original image.  Notice that the temperature decreases with increasing
time, a result of decreasing surface radiation and high net heat flux out of the ocean.

   T  3.  Escape and re-enter transect.  Move the right cursor to the bottom left of the image (the left
cursor should already be at the top left of the image).  Move both cursors to the right 100 pixels and
graph the vertical temperature profile.  As in the chlorophyll profiles, the surface is on the left with
depth increasing to the right.  Note that the surface temperature changes slowly in the upper 30-40m
of the water column.  Reposition the top cursor down 80 pixels.  This is the same depth as the top
of the deep chlorophyll maximum (DCM) you plotted earlier.  Move this cursor down an additional
20 pixels.  

Questions:
B1.  Over what range of depths is the largest temperature change, that is, where is the temperature
gradient highest?
B2.  The depth of the highest gradient is near the seasonal thermocline, a source of nutrientrich
water.  How many meters above or below this depth is the peak chlorophyll value?  

4.  Repeat step 9 from the chlorophyll image.  Draw the transect.  

Question:
B3.  Does the DCM trend follow the isotherm trend over time?  (Hint: count the number of isotherms
the transect crosses.)

   S ' C.  Current velocity image
   T ( 1.  Load FALVV56D and the VV LUT then display the image.  The speed of the north-south current
   T ^) component ranges from -80 cmsé1 (red = 60), which is directed to the south, to +80 cmsé1 (green =   ^)        p->*>*BB     T    196), which is directed to the north.  Blue (integer value of 128) is 05 cmsé1, the lower limit of the
sensor.  

   T  2.  Enter the transect option.  Move the topmost cursor to the top left.  Move the other cursor to the
top left and to the right 500 pixels.  Draw the graph at this depth (10m).  This velocity time series
illustrates mesoscale processes associated with a cyclonic eddy from day 1-15.  The mesoscale feature
has high velocities which are particularly evident in the mixed layer.  After the eddy passes the site,
the mooring is surrounded by Sargasso Sea water which has greatly reduced (blue) velocities (days
15-31).  

Question:
C1.  Do the temperature and chlorophyll data show the same mesoscale trends?  

3.  The first 15 days of the graph (step 2 above) also reveal a high frequency velocity variability. 
To determine if the higher frequency velocities have an inertial periodicity, the result of Coriolis
   T  effect, we plot an 11 day time series. Escape and re-enter transect.  Move the right cursor to the left
of the image, then 100 pixels to the right.  The other cursor should already be at the top left.  Plot
the time series.  Note the increase in velocity with time.

Questions:
C2.  What is the dominant frequency (in cycles/day)?  
C3.  The inertial period at this site is 1 cycle/22 hr (ie period = 22hr/day).  Is the periodicity of
velocity similar to the inertial period?  
C4.  Are the same motions evident for isotherms below the thermocline?

4.  Move both cursors down 20 pixels and re-graph.  

Questions:
C5.  Do the magnitudes of the velocities in 3 and 4 differ?  

When velocity layers do not differ in speed it is likely that the depth of the layer is greater than the
mixed layer depth.  Try to find the base of the mixed layer.  To do this, use the 10m time series as
a reference and re-plot velocities for other depths in 20 pixels increments.  For each plot it is best
   T  to escape to the main menu, re-enter transect, move the cursors to the top (10m) and draw the 10m
graph; then move the cursor to the next depth and plot that time series.  In this way, the reference
depth (10m) is always depicted as a blue line and the incremented depth as a yellow line.  

C6.  Is the base of the mixed layer at the same depth as the top of the thermocline: if not, where is
it?
C7.  Is the base of the mixed layer at the same depth as the top of the DCM: if not, where is it?  
C8.  Does the mixed layer depth change over time?


   S K$ D.  All three images
   T #% 1.  Use the Toggle function in the Multiimage option to compare each dataset with each of the
   T % others.  Note that you will need to create a stretch using Modify LUT which can be used for all three
images.  Using this, and the answers to the earlier questions, review to what extent the uses of image
analysis techniques using BILKO improves the analysis, interpretation and intercomparison of multiparameter datasets.
   ])         p->*>*BB  Ԍ   T    ԙaC References ă

For an introduction to moored, bio-optical measurements see:

Dickey, T.D., J. Marra, T.C. Granata, C. Langdon, M. Hamilton,  J.L. Wiggert, D. Siegel, and A.
Bratkovich. 1991. Concurrent high resolution bio-optical and physical time series observations in the
   T  Sargasso Sea during the spring of 1987. J. Geophys. Res., 96, 8643-63.  

Dickey, T.D., T.C. Granata, J. Marra., C. Langdon., J. Wiggert, Z. Chai-Jochner, M. Hamilton,
J. Vazquez, M. Stramska, R.  Bidigare, and D. Siegel. 1993. Seasonal variability of bio-optical and
   T s physical properties in the Sargasso Sea.  J.Geophys. Res., 98, 856898.

For an introduction to eddies and mesoscale processes see:

Robinson, A. R.  1983.  Eddies in Marine Science, Springer-Verlag, Berlin. 609 p.

Woods, J. 1988. Scale upwelling and primary production.  In: Towards a Theory on
   T ] Biological-Physical Interactions in the World Oceans, Ed. B.J. Rothschild, NATO ASI Series, Series
   T 7 C: Math. and Phys. Sci., 239: 7-38.


y   b                     d       d d                                                                         y 

This lesson was prepared by Timothy C. Granata, August 1992.

   R ! Comments or further suggestions should be sent to:

Dr. Timothy C. Granata
Ocean Physics Group
Department of Geological Sciences
University of Southern California
Los Angeles, CA 90089-0740
U.S.A.

FAX: (1) 213-740-8801
Email: Granata%sei@ramoth.usc.edu

   A         p->*>*BB  














































   S ' 
a7 Figure 1: The Biowatt Study Regionă
   (	         p->*>*BB     T    a> Answers for Lesson 1ă


A1.  The depth of the DCM peak is approximately 80m.

A2.  The DCM extends from 60-105m.  The base of the DCM is coincident with the 1% light level
   T  (110m and 10 Em2sé1).

A3.  The profile is not constant for time scales of less than approximately 3 days, thus, steady state
is a relative term.

A4-5.  Chlorophyll has 10-11 cycles or a periodicity of approximately 1/day, which coincides with
daily light periodicity.  Note the cycle on the sixth day is obscured by rising chlorophyll values
associated with a cyclonic eddy that is advected past the mooring.  The variability on shorter time
scales results from waves (eg. tides and internal), as well as physiological processes associated with
pigments.

A6.  The rate at which the DCM deepens is approximately 1m/day, which is the slope of the transect
line.  

B1.  The largest gradient is between 45-60m and is the seasonal thermocline.  

B2.  The peak DCM value is at approximately 80m, so it is 20m below the bottom of the seasonal
thermocline: it is associated with relatively high nitrate concentrations.  The chlorophyll in the DCM
may represent new production, that is, production based on vertical nitrate dispersion through the
thermocline.

   T  B3.  The DCM traverses three or four isotherms and does not seem to follow the general trend of in   T  situ temperature.  The deepening of the DCM may be the result of other advective processes, as well
as differential growth of cells and sinking of phytoplankton, marine snow, and detritus.  

C1.  Mesoscale activity is evident in the isotherm time series and is visualized as shallowing then
deepening isotherms during the passage of a cyclonic eddy.  This is most likely a result of the edge
of the eddy, composed of warm Gulf Stream water, passing over the mooring.  The DCM also
responds to mesoscale forcing as is evident by the coincidence of the highest chlorophyll values
during the passage of the eddy. Note that cyclonic eddies are upwelling structures that often have high
chlorophyll values associated with changing vorticity (or rotation) of the flow field.  

C2-3.  Current speed shows 12 peaks in 11 days, so a period of approximately 22 hr, the inertial
period.

C4.  The currents and isotherms both have an inertial period.

C5.  Very little difference is observed between velocities at 10m and 22m.  

C6.  The bottom of the mixed layer is between 34-45m, where the velocities decrease significantly
as a function of depth.  Thus, the mixed layer depth is near the top of the thermocline.  

C7.  The top of the DCM corresponds to the bottom of the thermocline and lies at least 25m below
the mixed layer (at 60m).   ])
        p->*>*BB  ԌC8.  The mixed layer changes as different water masses are advected past the mooring.  Note that
as velocity (kinetic energy) increases towards the end of the time series, the mixed layer deepens and
both the thermocline and the DCM move up in the water column.  


   8         p->*>*BB    r         p       QF      z N    # c     P E37ApP# `"l Practical Lessons: 2p               p-@*@*@@  
a. Practical Lessons Using Marine and Coastal Image Data


   T ` ua. 2: SPATIOTEMPORAL EVOLUTION OF THE COASTAL UPWELLING
	a3 IN THE SENEGALESE AND MAURITANIAN LITTORAL



   T   aB Aim of Lessonă

To show the dynamics of sea surface temperatures in an area of coastal upwelling along the
Senegalese coast of West Africa using Meteosat 4 infrared images.



   T 
 aC Objectivesă

After completing this lesson, you should be able to:

1) identify and characterize thermal structure along a seasonal coastal upwelling

2) relate the areas of low temperature to the bathymetry

3) synthesise the thermal information so as to characterize the strength and extent of the upwellings.



   T  ~a= Background Informationă

For this lesson you require a PC running the UNESCO processing toolkit, BILKO.  Some experience
in the use of BILKO to display satellite images is assumed.

   S  Image Data

The Meteosat images used to obtain SST (sea surface temperature) data come from the infrared sensor
of the Meteosat 4 geostationary satellite.  The daily images used are each compiled from 24 images
taken hourly, and extracted from the whole disc of Meteosat.  They have been linearly calibrated
using "in situ" temperatures measured by a network of merchant ships.

The relation between the temperature T, expressed in celsius and the pixel value (numerical count,
N) is:	 1     d d d d d d d d (1)                               1     d d d d d d d d (1)           !                   	
`!  #b)      d d d d d         d d                                                     
   b               T~=~55~~N over 2&m     P E37&P&m     P E37&P&m     P E37&P       T     }4N                      55      @ 2     e   `$  ""t#""!! "$  0    
         p-@*@*@@ )",  ! 
 0  ԒNote the negative gradient, so that higher numbers correspond to cooler temperatures.  Seven images
are used.  Each is the monthly mean of the daily temperature images from December 1991 to June
1992.


   S 8 /a; Table 1: Image descriptionă

# c     P 7ApP# m  
 ! d d x                       
 A Ed d x        pr<                    m   h		
p 
 
                 y O  	 File name   X         y O  
- Satellite   X         y O  x> Periods   X         y O  JwO Channel   X         y O  i\ Size   X         y O  l Resolutionă p 
 
P 	 	              S1291.DAT          Meteosat 4          131 Dec 91          $wS IR          256 X 256          ho 6 km P 	 	P 	 	X              S0192.DAT   	       Meteosat 4   	       131 Jan 92   	       $wS IR   	       256 X 256   	       ho 6 km P 	 	P 	 	              S0292.DAT   H       Meteosat 4   H       129 Feb 92   H       $wS IR   H       256 X 256   H       ho 6 km P 	 	P 	 		              S0392.DAT          Meteosat 4          131 Mar 92          $wS IR          256 X 256          ho 6 km P 	 	P 	 	H              S0492.DAT   
       Meteosat 4   
       130 Apr 92   
       $wS IR   
       256 X 256   
       ho 6 km P 	 	P 	 	              S0592.DAT   8       Meteosat 4   8       131 May 92   8       $wS IR   8       256 X 256   8       ho 6 km P 	 			
              S0692.DAT          Meteosat 4          130 Jun 92          $wS IR          256 X 256          ho 6 km 		 8   # &m     P 7b&P# 

   S h Oceanographic background information

This lesson concerns a 1500 km stretch of the West African coastline, from Northern Mauritania (21$
North) to Southern Guinea (9$ North), a coastline of some 1500 km in extent (Figure 1).  In this
area, the north trade winds are seasonal and blow from roughly November to May, which
corresponds to the dry, cold season as opposed to the wet, warm season which runs from June to
October.  The trade winds along these coasts are from North to South and generate an upwelling of
cold waters, a phenomenon called "coastal upwelling".  These cold waters are of high nutritive value. 
Upwelling areas are shown hatched on Figure 1, and the arrows indicate typical ocean surface current
patterns.

In general, the continental shelf is narrow North of Cap Vert (14$40- North) and becomes 
progressively wider to the south.  The seas to the south of Cap Vert may be treated as  belonging to
a separate ecological system.

The period of maximum upwelling corresponds to the period of maximum wind strength.  This is to
be expected since, according to Ekman theory, the rate of upwelling is physically linked to the
strength of the wind component blowing parallel to the continental slope.

A measure of the strength of upwelling is given by the Coastal Upwelling Index (CUI) which may
be calculated from the Ekman theory:

A  #b8+      d d d d d       &  d d                                                        b        ]     7 CUI ~ =~ { r ~C sub d ~V sup 2}over{2~ OMEGA~ sin phi }&m     P E37&P&m     P E37&P&m     P E37&P       CUI     `r     _`Cz z   &d     l`V       X     t z z    2      @ 2      @ sin      8@ 2      @ -] $  ""$""!A "$  0            p-@*@*@@ 8+".  A  0  Ԓ
 r:`	`	# air density
   T   Cd:`	`	# air/sea drag coefficient
   T   V2:`	`	# square of the wind component parellel to the coast
 :`	`	# angular velocity of the earth
 -:`	`	# latitude


   T  aA Lesson Outlineă


   T I	 1. Enter the Bilko software.  Load and display image S0392.  Keep the 16 grey level default palette
of the software and enhance the image contrast using only the Modify LUT option to distinguish the
   T 
 thermal structures over the sea.  For example, use the option Histogram equalisation.  Remember
that for the images in this lesson, a dark tone corresponds to higher temperatures.

Compare the image you obtain with Figure 1.  Note the nature and location of any remarkable
features.

   T  2. Load the palette SST.PAL and the look up table SST.STR, provided with this lesson.  The lowest
   T  temperatures are now shown in light blue and the highest in red.  With the Cursor option, determine
the minimum temperatures along three transects running from west to east (see Figure 2):
        1) just south of Dakar
        2) offshore from the Saloum estuary
        3) offshore from the Casamance river

On this image, can you estimate the southern limit of the thermal front of the upwelling?

   T  3. Display successively the monthly mean images of SST from December 91 to June 92 (S1291 to
   T ~ S0692) and determine visually the extent of the upwelling during this season. To investigate the
   T W change between successive months, use the Toggle option.  Describe what you see.

4. Calculate the mean lowest temperature off the south coast of Senegal on the three transects defined
in step 2, and plot these mean monthly values.  You will find that they approximate the values
reported in Figure 3 which represent the mean temperature during the 91-92 cold season.  Also shown
are the mean values from 1984 to 1988.  Comment on whether any climatological change is implied
by these curves.

5. Figure 4 shows the time series of an upwelling index theoretically calculated from the strength of
the wind component parallel to the coast, from January to May 1992 (see Background Information). 
How do you interpret this information in relation to the anomalous SST identified in step 4?



   S (% aC Referencesă

Coastal Upwelling (Coastal and Estuarine Sciences 1), F.A.Richards Ed., Washington, 1981

Rebert J.P., 1979, Aper'u sur l'hydrologie du plateau continental ouest-africain de la Mauritannie !
   T `) la Guin)e. In: Rapport du Groupe de travail ad hoc sur les poissone p)lagiques  c=tiers   `)        p-@*@*@@     T    ouest-africains de la Mauritanie au Lib)ria. COPACE/PACE S)r. 78/110, 165pp.


      y    b                     d       d d                                                                         y 

This lesson was prepared by H. Demarcq, E.S. Diop, J. Citeau & A. Pezin.


   R  Comments or suggestions relating to this lesson should be sent to:


H.Demarcq, J.Citeau, A.Pezin
CRODT/ORSTOM/UTIS
P.O. Box 2241 Dakar
Senegal

Telex: c/o ORSTOM:51468SG
Fax: (221) 34 27 92

E.S.Diop, Professor
c/o UNESCO
P.O. Box 3311
12, Avenue Roume Dakar
Senegal

Telex: 51410
Fax: (221) 23 83 93


            p-@*@*@@  










































   T H$ a/ Figure 1: Sea Surface Currents and Upwelling Areas
   S !% a3 During the Cold Season (February to April)ă
Aa? (From Rebert 1979a)   %         p-@*@*@@  










































   T H$ a= Figure 2: Area of Study
   !%         p-@*@*@@  


















   T  <a3 Figure 3: Sea Surface Temperature Evolution



















aaH 





   S % eaH Figure 4: Upwelling Index During the Cold Season in Dakar Station   %         p-@*@*@@  
   T   a> Answers for Lesson 2 

   T  1. Image S0392 shows the beginning of the coastal upwelling along the coast of Senegal and
Mauritania (Figure 1).  North of Dakar, the coldest waters are close to the coast because the
continental shelf is narrow.  The maximum upwelling of water occurs above the 200 m isobath, that
is, near the boundary of the continental slope.  South of Dakar, the coldest water is progressively
more distant from the coast because of the wider extent of the continental shelf.  Notice the filament
of cold water far offshore from the continental shelf to the west.

2. The mean of the three minimum temperatures you should find is approximately 19.5$C.  The
southern limit of the thermal front is located south of the Bissagos archipelago (Guinea Bissau).

3. Comparison of the seven images shows a considerable reduction in the surface temperature of the
coastal water from December to March.  From March to June, an increase of the temperature can be
seen.

4. Relative to the five-year climatic mean based on 198488, it may readily be seen that the mean SST
values of the 91-92 cold season are warmer from December to March (positive anomaly).  In March,
the mean SST corresponds approximately to the fiveyear mean and becomes progressively colder
from March to May (negative anomaly).  The anomaly disappears in June.  This corresponds
approximately to a onemonth delay of the cold season, with respect to the fiveyear mean.

5. During the 91-92 cold season, the shift observed in the maximum intensity of the upwelling
corresponds to a delay in the trade winds which are maximum in April in the Dakar area, instead of
the usual February-March.  Note that the wind is recorded at a single point (Dakar) which may not
be representative of the whole region between Dakar and South Senegal revealed by the satellite data. 
This may cause a slight mismatch of phase between the maximum wind and the minimum SST.


  QF        p       h    v  z N    # c     P E37ApP# `"l Practical Lessons: 3p     R        p-@*@*@@  # &m     P 7b&P# 
a. Practical Lessons Using Marine and Coastal Image Data
aaH 

   T ` ja+ 3: VARIATION OF AVHRRDERIVED SEA SURFACE TEMPERATURE WITH
#a<  SATELLITE ZENITH ANGLES



   T   aB Aim of Lessonă

To understand the scale and sense of the error induced by the satellite zenith angle on the calculation
of Sea Surface Temperature (SST), and the need to correct for it. 



   S 
 aC Objectivesă

On completion of this lesson, you should be able to:

1)  display VISE image data and convert locations given in pixel coordinates to ones in latitude and
longitude

2)  compare images of satellite derived SSTs in which any differences depend only on zenith angles

3)  detect the size and scale of those SST differences.



   T z ~a= Background Informationă

   S + Equipment

For this lesson you require a PC running the UNESCO image processing toolkit, BILKO (Version
1.3). To load the image display software, enter the directory in which the program is stored and type
<unesco13>.  No prior knowledge of the BILKO processing system is required.  This lesson
illustrates the application of some of the facilities provided by the toolkit.

   S   Other equipment

You may find it helpful to have a calculator.

   S s# Information about the images

The purpose of this lesson is to make a comparison of two satellite images in order to assess the error
induced by the zenith view angle.  For this, sensors on the NOAA 9 and 12 satellites were used.  We
first demonstrate the consistency of calibration, in channel 4, between these satellites for view angles
which are close.  We then compare the brightness temperature at different angles of view.

All the images were acquired at night, so the noise induced by solar heating of the ocean skin layer   [)         p-@*@*@@  in low wind conditions is eliminated.  The two satellite passes were recorded within 30 minutes.  As
the concentrations of minor constituents and water vapour content should not have changed
significantly we assume that atmospheric perturbations of the ocean radiation did not alter between
the images.  Similarly, the sea surface temperature is considered to remain constant from one image
to the other. 

Given these assumptions, it is not necessary to compare true temperatures.  The aim of the lesson is
satisfied by investigation of the influence of view angle.  The origin of this influence is the change,
with view angle, of the distance covered by the radiation in the atmosphere and of the emissivity of
the ocean.  At larger viewing angles, the distance covered increases and the emissivity decreases.

For this lesson, thermal infrared data values were converted to brightness temperature values using
the inverse of Planck's radiation equation.  Calibration coefficients were calculated from
measurements of the onboard blackbody and the deep space blackbody.  A geometric correction was
applied and is derived from knowledge of the characteristics of the sensors.  The location was refined
   T  manually (Fusco et al., 1990).  Data have been resampled onto a chart projection with a fixed
resolution of one nautical mile; the land has been masked.

A cloud mask was applied to the images.  This takes the form of a simple threshold on the thermal
channels 4 and 5.  To ensure that the brightness temperature calculated was not cloud contaminated
pixels displaying temperatures less than 20 C were masked.  This threshold is high enough to ensure
few erroneous categorisations.  A local uniformity test on a 2 ' 2 pixel array of the brightness
temperature of channel 5 was applied.

   S B Image Data

   T  Two 512 ' 512 images are provided for this lesson: VISE1.DAT and VISE2.DAT.  They were 
processed in the same way.  

# c     P 7ApP# | 
 A Ed d x        pr<                   
 a d d x     {   y}}L                          | & 		
 8 
  
 8&              Image
name          Date          Latitude          Longitude          Equator Crossing
Time         Satellite Zenith Angle& 8 
  
 P  	 	{&                                                      NOAA9          NOAA12          NOAA9          NOAA12" P  	 	P 	 	"              VISE1   S       24/06/92   S       11.615.9$S   S       4549.25$E   S       1601   S       1634   S       1845$   S       1218$ P 	 			              VISE2          03/09/92          20.825.3$S          40.544.76$E          1635          1654          021$          3346$ 		 S   # &m     P 7b&P# 
Each is composed of four 256 ' 256 images representing the same geographical area (see the
diagram on the next page).  These NOAA images were acquired at the SEAS station (La Reunion),
which archives daily HRPT and TOVS images of the Indian Ocean.

The full resolution is 1 nautical mile ' 1 nautical mile, so 60 pixels cover one degree of latitude and
   T # of longitude. The images to the left are projections of channel 4 brightness temperature (Tb4)
calculated, respectively, from the data of the NOAA 9 and 12 satellites.  The images to the right are
the values of the zenith view angle () for each pixel.  This is the angle subtended between the zenith
and the sensor for an observer on the ground at the location of the pixel.
	 1     d d d d d d d d (1)           A                  1         d d     d d (1)           !                   	

   (        p-@*@*@@  Ԍh  
 a d d x     {   y}}L                        
  $/ / / /        S'        h  	V 
S                j J _  1`mHE 256ă

   S  < NOAA 9
   S  n> Tb4

  j J s 8 line 256   V        j J _  257`d?X 512ă

   S  O NOAA 9
R  V 
	                

]< NOAA 12
   S  n> Tb4ă

  j J 	 ; line 512   
      

   S E rO NOAA 12
   S  R ă 	 V   
   S  
a7 Content of VISE1.DAT and VISE2.DATă

The view angles are given directly in degrees

   T  The temperature precision is 0.1$C.  The pixel value, CN, is related to the brightness temperature,
   T  Tb, by:
za  #b      d     d d         d d                                                        b               T sub b~=~16~+~CN over 10&m     P E37&P&m     P E37&P&m     P E37&P       Tz z     b     4CN                     16      @ 10        z$  """"`!a "$ 
Two further image files are provided.  They are both derived from VISE1 data and have been
produced as 256 ' 512 images.

   T  VISE3.DAT shows on the left the NOAA 12 brightness temperatures subtracted from the NOAA 9
temperatures: the values on the screen are absolute differences shown with a resolution of 0.1 degC. 
On the right, the absolute differences in zenith angle are shown with a resolution of 1$.

   T  VISE4.DAT shows the absolute difference of the secant of the zenith angle,  sec .  The pixel
values shown on the screen are the differences multiplied by 50.



   T & aA Lesson Outlineă


   T ! 1.  Use the load image function.  Notice that the maximum dimension you can display is 512 by 256. 
   T " Start by loading the first 256 lines of VISE1; then look at the image using display image.

   T 9$ 2.  You have to load the LUT VISE1 and the palette VISE1 to augment the contrast and facilitate
the visualisation.

   T & 3.  First, you need to learn to interpret these images.  Use the cursor function and, for a few pixels
try to establish a correspondence between temperature and view angle.  For example, the temperature
given in column 50 (x coordinate = 50) corresponds to the same location on the ground as the pixel
with view angle given in column 50 + 256 = 306.  Try also to convert the coordinates of a pixel 0  K)        p-@*@*@@ "  a  0  into latitude and longitude.  (If you display in full resolution, remember that 60 pixels represent one
degree.)

The location for the pixel at column 50 and line 100 is (VISE1.dat at full resolution):
 `	`	# , lat = (20.83 + 50/60)
 `	`	# , long = 40.5 + 100/60

Note that clouds are assigned the value 250.

   T  4.  Display the complete image.  You have to sample the lines and the column (2).  Now the
resolution is 30 pixels for one degree.  You now have to add only 128 to the temperature pixel
column number to find the location of its view angle information.

5.  Now you can compare data from the two satellites.  First, you have to ensure that the consistency
of the satellite calibrations is good.  The surface temperature and the signal attenuation should be
approximately the same in the two images for pixels which correspond to the same angles of view. 

6.  Find an area that has the same view angle for the two satellites.  What is approximately the zone
in which this occurs?

Once you have determined the area, compare the brightness temperature and note the difference for
a few selected pixels.  (Do not forget that you have to divide by ten to obtain temperature difference
in degrees).  Estimate the mean difference for the pixels chosen.

7.  Find the areas where the brightness temperatures differ from one image to the other and compare
the angles.  Estimate the general variation.

   T  8.  Display VISE2: load the palette VISE2 and the LUT VISE2.

You should notice that as a consequence of the temperature error caused by view angle error, some
clear pixels are treated as cloud pixels by the temperature test.

The difference of angle is about the same throughout the area between the two passes.  What is its
value?  Evaluate the mean difference temperature for this difference angle.

   T  9.  Display VISE3, the associated LUT and palette.  This image shows the absolute differences of
temperature and angle for the two passes of the first image.  Compare this result with your previous
answer (question 7).  You can notice that the angle difference is only 30$ in the lower part of the
image (less than the upper) and the temperature difference is higher (about one degree).  The error
   T   induced by the satellite zenith angle is not linear.  To confirm this, display VISE4 (with the same lut
and palette): now the brightness temperature difference between overpasses (shown on the left) is
compared with  sec  (on the right).  Observe that the sense of variation is now the same.  Use the
cursor function to confirm this for a few selected pixels.

If you wish to study further the effect of the zenith angle  on the multichannel atmospheric
   T % correction of seasurface temperature data, you should refer to the NOAA Polar orbiter data  User
   T & guide and the papers by Antoine et al (1992) and Barton (1989).  All the algorithms have their own
   T ' coefficients, found by regression with in situ data or with a model of radiative transfer.  They show
how difficult it is to obtain a correction which is applicable everywhere.
   b)         p-@*@*@@  Ԍ# &m     P 7b&P#   
   T   aC Referencesă

For information on the Sea Surface Temperature and NOAA satellites:

   T 9 NOAA polar orbiter data.  Users guide.  U.S. Department of Commerce.  December 1991.

Antoine, J.Y., M.Derrien, L.Harang, P.le Borgne, H.le Gleau, C.le Goas, "Errors at large zenith
   T  angles on AVHRR derived sea surface temperature".  Int. J. Remote Sensing, 1992, 13, (9), 17971804.

   T M	 Barton, I. Comparison and optimisation of AVHRR Sea Surface Temperature Algorithms.  Journal
   T '
 of Atmospheric and Oceanic Technology, 6, December 1989.

   T  Fusco, L., K.Muirhead, G.Tobiss.  Earthnet's coordination scheme for AVHRR data.  Int. J. Remote
   T  Sensing, 1989, 10, (4 and 5), 625636.

   T e Dalu,G. Emittance effect on the remotely sensed SST. Int. J. Remote Sensing, 1985, 6, (5), 733740.

Masuda, K., T.Takashima, Y.Takayama.  Emissivity of pure and sea waters for the model sea surface
   T  in the infrared window regions.  Remote Sensing of Environment, 1988, 24: 313329.

Saunders, R.  An improved method for detecting clear sky and cloudy radiances from AVHRR data. 
   T y Int. J. of Remote Sensing, 1988, 9, (1), 123150.


      y    b C                    d       d d                                                                         y 

This lesson was prepared by P. Lena, M. Petit, M. RakotaRavalonstalama and M. Slepoukha.


Comments or suggestions relating to this lesson should be sent to:

P.Lena
ORSTOM / Station SEAS
BP 172  97492 Sainte Clotilde Cedex, La Reunion
France

email: lena@orstom.orstom.fr (internet, uunet)

Dr. M.Petit
ORSTOM / Station SEAS

email: petit@orstom.orstom.fr (internet, uunet)
 M.petit (sciencenet, omnet)

M.RakotoRavalonstalama
   R & ORSTOM / Station SEAS

   R ( email: rakota@orstom.orstom.fr   (         p-@*@*@@  ԌM.Slepoukha
   R   ORSTOM / Station SEAS

   R  email: slepoukha@orstom.orstom.fr            p-@*@*@@  
   T   a> Answers for Lesson 3ă

6.1  The upper right of the images has about the same view angle.

6.2  The difference is less than 0.4 degree and about 0.1  0.2 when we calculate the average.  So,
we consider that the calibration between the two sensors is good.  (The noise of the sensors is
estimated at 0.1$)

7.   The answer is found by viewing image VISE3.dat (see question 9).

8.1  30 degrees

8.2  0.6 $C

9.
   
         p-@*@*@@                p-@*@*@@  # &m     P 7b&P#   h        p           ~N  z P    # c     P E37ApP# `"l Practical Lessons: 4p 
a. Practical Lessons Using Marine and Coastal Image Data


   T ` a7 4: SEASONAL AND YEARLY VARIATION OF
   S 9 3a/ SURFACE TEMPERATURE DISTRIBUTIONS IN THE NORTH SEAă



   T   aB Aim of Lessonă

To understand the thermal structure of the North Sea by showing the seasonal course and the annual
variability of the sea surface temperatures (SSTs), as measured by NOAA satellites.



   T 
 aC Objectivesă

On completion of this lesson, you should be able to:

1) recognize the general structure of the seasonal course of SSTs in the North Sea

2) recognize the general circulation pattern of the North Sea

3) interpret the influence of differences in water depth on the SSTs

4) recognize possible significant deviations of the observed SST from the climatology.



   T S ~a= Background Informationă

   S  The BILKO image processing software

   T  Use BILKO version 1.3 (or later).  No prior knowledge of BILKO is necessary. Press <H> for help
   T  and <ESC> to return to the main menu.


   S   Image loop display program RUNPRES

RUNPRES is a program which displays a sequence of images (RUNPRES is produced by Advanced
   T " Graphics Software, Inc., USA).  This sequence is defined in file SSTLOOP.SHO. The SST*.PCX
   T u# files are screen dumps of BILKOdisplays of SSTimages.  The .PCXĩformat is compatible with
   T N$ programs such as SlideWrite and PC Paintbrush and is therefore also accessible with WordPerfect. 
RUNPRES needs an EGA (256K graphics memory) or VGA graphics card (VGA offers a better
performance of the display of the SST image loop).

The program automatically displays the image sequence, showing each image for approximately 12
seconds, depending on whether you run from diskette or from the hard disk.  The fastest way to
   T ^) display the loop is to copy RUNPRES.EXE, SSTLOOP.SHO and all .PCXĩfiles to your hard disk   ^)        p->*>*BB  and start the program from there.


   S  Other equipment

To work through the last part of this lesson you should obtain a calculator with sine and cosine
functions.


   S  Image data

This lesson is based on a very large number of images of the North Sea, received from the NOAAsatellites at the Royal Netherlands Meteorological Institute (KNMI).

The KNMI operates a groundstation for the reception of raw NOAAAVHRR imagery (in HRPT
format).  Image data of all NOAA orbits passing over the North Sea are received and processed at
KNMI in real time.  Atmospherically corrected SST images are derived by combining the 11m and
12 m AVHRR bands (channels 4 and 5), using the socalled splitwindow technique.  Coefficients,
derived by the Rutherford Appleton Laboratory, are used in the KNMI processing scheme (see
   T  LlewellynJones et al. 1984).  The accuracy of the NOAA SSTs can be trusted to within  0.6 $C.

It is essential to understand that the satellite measures the radiance emitted by the surface of the sea
and thus the temperature of the sea surface.  How closely this surface temperature is related to the
temperature of deeper water depends on how well the sea water is mixed in the vertical direction. 
Under certain weather conditions, low wind speeds and strong solar radiation, the sea surface can be
several degrees warmer than the deeper water.  Under the circumstances, local warm areas ("hot
spots") may be visible in the SST images.
 
Since January 1990, KNMI has maintained a service which produces weekly composite SST maps of
the North Sea.  The weekly SST maps are derived from the complete set of AVHRR data received
from all NOAA satellites in the course of one week.  The "maximum value" technique is used for
the composition of the weekly SST map.  Therefore any images showing obvious "hot spot" effects
have been left out of the compositing process.

The North Sea has a very cloudrich climate.  An SST image of one single orbit is, in most cases,
heavily cloudcovered.  The composite technique can be helpful in obtaining a less cloudcovered
image.  However for many weeks during the year even the weekly composites show large areas
obscured by clouds, although images from more than 30 orbits contribute to the weekly composites. 
For the purpose of this lesson monthly composites are used to show the seasonal course and the yearly
variations of the SST distribution in the North Sea.  The monthly composites are derived by
   T ! averaging the weekly composites.  Thus every image shows the average SSTdistribution in the North
Sea for a period of one month and contains data from more than 120 NOAA passes. 

The seasonal course of the SST distribution in the North Sea is shown by looping a sequence of the
monthly compositeimages from 1990 and 1991 (SST*.PCX files).  To make the SSTloop look
smoother, extra images, derived by interpolation, are inserted between the monthly composites
(SST*A.PCX files).

Four NOAA AVHRR image files in BILKO format, covering a winter and a summer situation, are
provided for this lesson which demonstrates the annual variability of the North Sea SSTs (the .PCX   [)        p->*>*BB  files cannot be processed with the BILKOsoftware):


   T   a; Table 1: Image Descriptionă

J 
  $/ / / /        S'      
  Vd d x     9   o        J   
 
9                 S  E
 Image Filename             S  S Descriptionă  
` 	9              
# SST0290.DAT          I SSTcomposite of February 1990 ` 	` 	              
# SST0291.DAT   y       I SSTcomposite of February 1991 ` 	` 	              
# SST0890.DAT   	       J SSTcomposite of August 1990 ` 		y              
# SST0891.DAT   i       J SSTcomposite of August 1991 	 	   

All images are polar stereographically projected at a reference latitude of 52$N.  They cover an area
between 59$32.2-N, 2$56.6-W (Northwest corner) and 49$42.9-N, 9$28.2-E (Southeast corner). 
Each image contains 283 pixels ' 256 lines, which are derived from original full resolution SST
images by selecting every 4th line and every 3rd pixel.

   T Q Geographical coordinates Lat and Lon (in decimal degrees) can be derived from pixel position (X,Y)
using the following formulae:
  #bE      d     d d       9  d d                                                        b        x     x Lat~=~90~~2~arctan~{ LEFT ( SQRT{ { (3X~~159.3) sup
2~+~(4Y~+~3098.7) sup 2 } } over {11395.5} RIGHT ) }
~~~~~~~(degN)&m     P E37&P&m     P E37&P&m     P E37&P       Lat     4X     6
4Y      degN      Z               4      4      
4       90       2       arctan      `4(      43      w	4159      
4.      
43      4)z z    R|2      4(      44      43098      4.      H47      4)z z    |2      
@ 11395      @ .      @ 5       (      w )      6h       j       i      C6o      C q      C p      
     `	      	Px $  """"8! "$   #b}      d     d d         d d                                                        b        9     T Lon~=~arctan ` LEFT ( {3X~~159.3} over {4Y~+~3098.7}  RIGHT 
) ~~~~~~~~~~~~~~(degE)&m     P E37&P&m     P E37&P&m     P E37&P       Lon     4X     @ Y     =
 degE             4      @       A arctan      43      x4159      4.      43      u@ 4      E@ 3098      @ .      @ 7       
 (       )       h       i      X o      X p	     ] 9 ߶$  "";""`! "$  X "
   T s Similarly, pixel position (X,Y) can be derived from the geographical coordinates Lat and Lon (in
decimal degrees) of a location using the following formulae:
  #bg#      d     d d       0  d d                                                        b              X~=~{1} OVER {3} ` LEFT LBRACE~ 159.3~+~ {11395.5~ sin ` (Lon)
~sin ` { LEFT ( {90~~Lat} OVER{ 2} RIGHT) }} OVER{cos ` {LEFT
(  {90~~Lat} OVER {2} RIGHT ) }} RIGHT RBRACE  &m     P E37&P&m     P E37&P&m     P E37&P      X     l	Lon     (1Lat     {4Lat                   j
1      
4            J      Jv      JR      J.      J
      J      J      J      Jz      J      J}      JY      J5      J      J       J       J       J       J       !      v$      R$      .$      
$      $      $      $      z$      "      }$      Y$      5$      $       $       $       $       $       #           h      i      6o      6p#     J      T	 h      T	 i       o       p#     	       =1      I3      159      .      3      11395      k.      5      Esin      /	(      
)      sin      b190      
=2      L cos      	490      
@ 2 ߃$  ""%""! "$   #bO)      d     d d       0  d d                                                        b              Y~= {1} OVER {4}  ~LEFT LBRACE~ 3098.7~+~{11395.5~cos `
(Lon)~ sin ` {LEFT ( {90~~Lat}OVER{2} RIGHT ) }} OVER {cos `
{ LEFT (  {90Lat}OVER{2} RIGHT ) }}RIGHT RBRACE &m     P E37&P&m     P E37&P&m     P E37&P      Y     6
Lon     1Lat     4Lat                   f      41      x4      O      ,      ,v      ,R      ,.      ,
      ,      ,      ,      ,z      ,      ,}      ,Y      ,5      ,      ,       ,       ,       ,       ,       Z!      Zv$      ZR$      Z.$      Z
$      Z$      Z$      Z$      Zz$      Z"      Z}$      ZY$      Z5$      Z$      Z $      Z $      Z $      Z $      Z #     3      h      i       o       p#     
      _
 h      _
 i       o       p     
       g=1      gI4      3098      .      7      K11395      .      E5      cos      	(      T)      sin      ,
190      \=2      W	 cos      
490      @ 2 ߏ$  ""
#""! "$  `            p->*>*BBA E"    }"     g#"(    O)".    `  Ԓ
The image data are scaled between 2.3 C and +23.0 C with a temperature resolution of 0.1 degC.
   T  Pixel value 0 is used for cloud obscured pixels, value 1 is land and other pixel values N can be
transformed into SSTs by:
 #b	      d     d d       c
   d d                                                                 R       SST~=~0.1~N~~2.5~~~~~~~~~~~ (C)&m     P E37&P&m     P E37&P&m     P E37&P      @ SST     w@ N     	@ C      @       A@       A@ 0      @ .      @ 1      @ 2      [@ .      @ 5      Z	@ (      
@ )R $  ""b""!"$ 
   T  Two lookup table files are provided, AUG.STR and FEB.STR, for use with the image data from
   T  corresponding months.  Two palettes, SSTVGA.PAL and SSTEGA.PAL are also supplied.


   S $
 Hydrography of the North Sea

For the interpretation of the satellite imagery some aspects of the hydrography of the North Sea will
be considered here.

In relation to the bathymetry, the North Sea can be divided into three parts (see Figure 1): the
Southern Bight (5154$N) with water depths generally less than 40 m; the central North Sea (5457$N) with water depths of 40100 m except for shallower areas on the Dogger bank and along the
western coastline of Denmark; and the northern North Sea (north of 57$N) which includes an area
of shelf water 100200 m deep and the Norwegian Channnel with water depths from 200m to more
than 700 m in the Skagerak between Denmark and Norway.

In winter most parts of the North Sea are well mixed in the vertical direction by tidal action and
winds.  In summer, when a seasonal thermocline extends over most of the central and northern North
Sea in response to solar heating, the combination of variations in water depth and in tidal currents
leads to the development of distinct hydrographic regimes.  The Southern Bight and the shallow
coastal waters are well mixed throughout the year.

Figure 2 shows the generalized nearsurface pattern of water movement in the North Sea.  Atlantic
water enters the North Sea through the English Channel and around the north of Scotland.  These
inflows, together with the effects of the Earth's rotation (Coriolis force), drive the cyclonic pattern
of circulation in the North Sea.  The Norwegian coastal current is a distinct dynamic feature
influenced by deep inflow from the Atlantic, surface outflow from the Baltic and the spring snowmelt, as well as by the general circulation of the North Sea.  In contrast, the currents in the central
North Sea are weak and variable in direction.

Figures 3 and 4 show the climatic SST maps of the North Sea for the months February and August. 
   T   The maps represent monthly average SSTs, based on 50 years of insitu measurements of SSTs.  All
figures are from Lee and Ramster (1981).


   S N$ The weather situation over the North Sea

   T % The weather is a factor which dominates the state of the North Sea surface (roughness and
temperature).  The dominating westerly wind direction during the whole year makes the North Sea
a factor which strongly influences the climate of the countries on the eastern side of the North Sea.

The North Sea weather is very unstable.  Depressions, originating from the Atlantic, frequently result 0  ^)          p->*>*BB 	"    0  in severe storms over the North Sea causing wave heights of more than 10 metres.  When pressure
is high, the North Sea may develop a smooth surface which is quickly warmed by solar radiation.

For the purposes of the second part of the lesson an outline of the weather situation over the North
Sea is required.  For the months of February and August 1990 and 1991 the following description
should suffice:

   T  February 1990: , A very stormy month.  Mainly westerly winds with average speed of 15 msé1. 
The average air temperature was very mild, 8 C"
   T  August 1990: `	`	#  X, With respect to the wind a fairly calm month: average wind speed 5 msé1. 
Average air temperature at the central North Sea was 15 C."
February 1991: , A strong stable high pressure system resulted in a calm and cold atmosphere
   T  
 over the North Sea: the average wind speed was 10 msé1.  The average air
temperature was 2 C in the first half and 6 C in the second half of the month."
August 1991: `	`	#  X, A normal weather situation according to the climatology: average wind speed
   T  10 msé1,, average air temperature 16 $C."

The term "climatology" is used to refer to average conditions for a given month, derived from many
years of data.



   T  aA Lesson Outlineă

aD Section Aă


1.  From the MSDOS prompt line (ie. outside the BILKO programme) type RUNPRES
   T  SSTLOOP.SHO <ENTER>.  A sequence of SST monthly composites between January 1990 and
December 1991 will be displayed. Between the monthly composites, interpolations are shown.

2.  Watch the loop for a while to form an impression of how the temperature distribution of the North
Sea alters in the course of the year. 

The same colour scale is used for all images in the SSTloop:

# c     P 7ApP#     
J 
  Vd d x     9   o      
  d d x     *!   V        J  p 
p 
*              
3 Colour (VGA mode)           #P SST range ($C) p 
P 	*              16 (black)   !       20.6  23.0 P 	P 	               15 (red)   :#       19.1  20.5 P 	P 	!              14 (light red)   $       17.6  19.0 P 	P 	:#              13 (orange)   %       16.1  17.5 P 	P 	$              12 (light orange)   *'       14.6  16.0 P 	P 	%              11 (yellow)   z(       13.1  14.5 P 	P 	*'              10 (yellowgreen)   )       11.6  13.0   )!        p->*>*BB   P 	P 	z(               9 (green)   P       10.1  11.5 P 	P 	                 8 (greenblue)           8.6  10.0 P 	P 	P               7 (light cyan)           7.1   8.5 P 	P 	               6 (cyan)   @        5.6   7.0 P 	P 	               5 (dark cyan)           4.1   5.5 P 	P 	@               4 (light blue)           2.6   4.0 P 	P 	               3 (blue)   0	        1.1   2.5 P 	P 	               2 (dark blue)   
       ܩ2.3   1.0 P 		0	               1 (green)             land 	 
   
# &m     P E37b&P# 
  Use the following keys to control the loop :
   T x  X <ESC> ,  X4 : stop RUNPRES"
   T P  X <HOME> ,  X4 : skip to first image (SST of January 1990)"
   T (  X <END> ,  X4 : skip to last image (SST of December 1991)"
   T    X <> X`	`	#  X,  X4 : show previous image"
   T   X <P> X`	`	#  X,  X4 : pause; hold image"
   T   X <any other key> , : continue with next image"


3.  Try to understand the changes of the SST distribution during the year. Consider how the air
temperature, solar heating, movement of water masses, and water depths influence the SST.  List the
processes involved and suggest how they may interact.

4.  In which months are the three main currents of the North Sea mostly recognizable in the SST
images?  Why particularly in those months?

Use Figure 2 to identify the three main currents: the inflow of Atlantic water through the English
Channel, around the north of Scotland along the east coast of the United Kingdom and the Norwegian
coastal current.

5.  Can you distinguish shallower and deeper areas in the SST images?  How?  Use Figure 1 to
compare the SST distribution with the bathymetry of the North Sea.

   T ! 6.  Exit RUNPRES by pressing the <ESC> key.


aD Section Bă


   T & 7.  Enter the BILKO program by typing UNESCO13 <ENTER>.

   T ' 8.  Choose Load palette and enter either SSTEGA or SSTVGA, depending on the type of graphic
card in your PC.   ("        p->*>*BB  Ԍ   T    ԙ9.  Choose the Load LUT option and enter AUG.

   T  10.  Choose Multi image and Toggle. Enter filenames SST0890 and SST0891. Toggle between
images by pressing the spacebar (image 1 is the SSTcomposite of August 1990 and image 2 of
August 1991).

The colours represent the same temperature range in both SST images: dark red > 20 C and dark
blue < 12.8 C; each colour in between covers a range of 0.6 degC.

First try again to recognize one or more main currents in each image.  Then compare images with
one another and with the SST map in Figure 4, which represents the climatology of SSTs in August. 
Notice that the patterns in both images correspond roughly to the map, but that the absolute SST
values may differ.  Where are the largest differences between both images and between the
climatology to be found?  Explain.

   T  11.  Choose the Load LUT option and enter FEB.

   T [ 12.  Perform exercise 10 again for the images SST0290 and SST0291 (SST composites of February
1990 and 1991).  FEB.LUT has been created in conjunction with the ? palette to have the following
interpretation: dark red > 10 C; dark blue < 1 C; each colour covers a range of 0.75 degC.  Figure
3 shows the SST climatology in February.

13.  To obtain a more quantitative picture of the annual SST variability in the North Sea extract
absolute SST values from the four images at locations A, B, C and D in Figures 3 and 4 by
performing the following steps:
 X ©  Calculate the pixel positions (X,Y) of the locations in the images (remember to use the
right dimension, degrees or radians, for input of Lat and Lon in the formulae) and write X,Y
coordinates in the table.  Use the formulae given in the Background Information."
   T   X ©  Load and Display image file SST0290."
   T }  X ©  Use the Cursor option to obtain the pixel value f(X,Y) at points A,B,C and D.  Convert
this to SST using the formula provided in the Background Information."
 X ©  Repeat the procedure for image files SST0291, SST0890, SST0891."
 X ©  Read the "climatological" SSTs at the four locations in February and August from Figures
3 and 4 and fill in the table."
   #         p->*>*BB  ^ 
  d d x     *!   V      
  d d x       $   n\n<                    ^   	
 
 
 

               Location          Lat,Lon          X,Y          Pixel value          SST ($C)          Climatology ($C)  
 
8 	 	                
 A          
57$N,
()1$W                    SST0290:
SST0291:                  February:" 8 	 	8  	 	"                           @          @       SST0890:
SST0891:                  August:" 8  	 	8 	 	"              
 B          
56$N,
   2$E   x          x       SST0290:
SST0291:   (         (      February:" 8 	 	8  	 	"                 D	          	          	       SST0890:
SST0891:   `
         `
      August:" 8  	 	8 	 	("              
 C   |       
54$N,
   7$E                    SST0290:
SST0291:                  February:" 8 	 	8  	 	`
"                 
                             SST0890:
SST0891:                  August:" 8  	 	8 	 	"              
 D          
52$N,
   3$E   X          X       SST0290:
SST0291:                  February:" 8 	 	h		"                 $                           SST0890:
SST0891:   p         p      August: h		    

14.  What can you learn from the numbers in the table about the annual variability of the SSTs in the
North Sea considering the hydrography at the locations and the actual weather situations (see
Background Information)?



   S 0 aC Referencesă

   T  Lee A.J. and Ramster J.W. (editors), 1981. Atlas of the seas around the British Isles.
M.A.F.F., Lowestoft, U.K. Her Majesty's Stationary Office Publication.

LlewellynJones D.T., Minnett P.J., Saunders R.W. and Zavody A.M., 1984. Satellite
multichannel infrared measurements of sea surface temperature of the N.E. Atlantic Ocean using
   T   AVHRR/2. Quarterly J. Roy. Met. Soc., 110, p. 613.


     " y    b (                    d       d d                                                        $                 y 

This lesson was prepared by R.W.M.E. Verhoeven and J.N. Roozekrans,  June 1992.
The realization of this lesson was cosponsored by the Netherlands Marine Research Foundation.
   <%$         p->*>*BB     R    Comments or suggestions relating to this lesson should be sent to:


Mr. J.N. Roozekrans
Royal Netherlands Meteorological Institute
P.O. Box 201
3730 AE De Bilt
The Netherlands

Fax: 31 30 21 04 07

   H	%         p->*>*BB  


















   S  Figure 1: Bathymetry of the North SeaF O Figure 2: Generalized nearsurface   {   {   {   {   {   {   {   { pattern
   T  of water movement in the 4 hh= F O ppX a   i xxr North Sea 






















   S I$ Figure 3: SST climatology of thehh= F O Figure 4: SST climatology of the 
   S !%  North Sea in February ($C)hh= F O North Sea in August ($C)   !%&         p->*>*BB  
   T   a> Answers for Lesson 4ă

3.  The SST can be changed by convective and advective processes.  The advective processes concern
the transportation of water by currents and tides.  The monthly composite images only reflect the
main currents in the North Sea.

The convective processes concern the airsea interaction, solar heating and vertical mixing of the
water column.  The sea surface can be heated by a warmer atmosphere or by solar radiation.  Winds
and tides cause vertical mixing in the water column and thus transportation of heat towards deeper
layers.
 X "
The water depth (bathymetry) has a large influence on the mechanism of cooling and warming of the
North Sea water as a result of airsea interaction and solar heating.  Deep waters have a larger heat
capacity than shallow waters.  The large buffer capacity of deep waters for storage of heat results in
a slower reaction to atmospheric and solar influences (heating up or cooling down) in deep waters,
relative to shallow waters.

4.  The visibility of the main currents in the SSTimagery is dependent on the contrast between the
temperature of the North Sea and the temperature of the water in the current.  This contrast is highly
dependent on the season.
 X i) The current through the English Channel:"
This current is an affluent of the eastern Atlantic Ocean. Throughout the seasons the temperature of
the East Atlantic water is much more stable than that of the North Sea.  The temperature contrast
between the Atlantic water and the North Sea is at a maximum during the winter months (November
until March), when the southern North Sea water is cooled more strongly than the Atlantic water. 
In some months (November and December) the current water can be traced through to the north of
Denmark.
 X ii) The current around the north of Scotland along the British east coast:"
This water is relatively cold at all seasons.  In the winter the temperature of the ocean water is a little
warmer than that of the North Sea.  In the summer the penetrating ocean water is much colder than
that of the North Sea and therefore is clearly visible in the SST images of June until August.
 X iii) The Norwegian coastal current:"
Surface water from the shallow waters of the East Sea, the Kattegat and the Skagerak streams along
the Norwegian coast in a northward direction.  In the winter this water is colder and in the summer
warmer than the North Sea so the Norwegian coastal current is recognizable in the SST imagery
during the periods JanuaryMarch and JuneAugust.

5.  Yes. As already explained in question 3, the bathymetry has a strong influence on the mechanism
of cooling and warming in the North Sea.
 X "
In the spring the shallow waters along the coasts become distinguishable as warmer areas, as may be
seen in the SST imagery for April when the radiation of the sun starts to heat up the North Sea. This
situation continues until September. In October the situation is reversed: the shallow waters are cooler
than the deeper waters.
 X "
The shallow Dogger Bank in the central North Sea is visible as a warmer area in the SST images of
AprilMay 1991 and as a colder area in December 1991.

10.  The cold current of ocean water along the British east coast is clearly visible in both images.    Y)'         p->*>*BB  The Norwegian coastal current is partly visible.

The SST image of August 1991 (image 2) fits better to the climatologymap than the image of 1990
(image 1).  The fact that the weather situation in August 1991 was normal might be the reason for
this.

In 1990 the shallow waters along the coasts in the southern part of the North Sea are a few degrees
warmer than in 1991 but the cold deep waters in the northwest part extend further to the south in
1990.  The lower wind speeds in 1990 do not explain this difference in the SST pattern.  An
explanation may be a difference in the distribution of solar heating over the North Sea in the two
years.

12.  The current through the English Channel is easily recognized, especially for February 1990.  The
Norwegian coastal current is just discernable.

The SSTs in February 1990 (image 1) were much higher than in 1991 (image 2). The air temperatures
in 1990 were very high with regard to the climatology. The extremely high winds contributed to a
rapid transfer of the atmospheric heat to the North Sea water.  The largest SST difference between
1990 and 1991 can be found in the shallow waters in which temperatures are more sensitive to
atmospheric variations.  The definition of the "cold" shallow Dogger Bank in the SST image for
February 1991 is remarkable.

13.  The completed table is shown below

r 
  d d x       $   n\n<                  
 d d x     @(   n<                    r " h		 
 
"              Location          Lat,Lon          X,Y          Pixel value          SST ($C)          Climatology ($C)  
 
8 	 	@              
 A          
57$N,
()1$W   H       
33,
  69   H       SST0290:  99
SST0291:  85         h\ 7.4
h\ 6.0         February:   6" 8 	 	8  	 	"                                            SST0890: 167
SST0891: 167   0      :[ 14.2
:[ 14.2   0      August:    13" 8  	 	8 	 	"              
 B   L       
56$N,
   2$E          
94,
  96          SST0290: 102
SST0291:  85   h      h\ 7.7
h\ 6.0   h      February:   6" 8 	 	8  	 	0"                                            SST0890: 182
SST0891: 186         :[ 15.7
:[ 16.1         August:    15" 8  	 	8 	 	h"              
 C          
54$N,
   7$E   (        
204,
  144   (        SST0290:  89
SST0291:  63          h\ 6.2
h\ 3.8          February:   3" 8 	 	8  	 	"                 !          `"          `"       SST0890: 212
SST0891: 201   #      :[ 18.7
:[ 17.6   #      August:    17" 8  	 	8 	 	 "              
 D   ,$       
52$N,
   3$E   $       
122,
  205   $       SST0290: 113
SST0291:  83   H%      h\ 8.8
h\ 5.8   H%      February:   5" 8 	 	h		#"                 d&          &          &       SST0890: 207
SST0891: 197   '      :[ 18.2
:[ 17.2   '      August:    17 h		 H%   
14.  Some of the conclusions are already discussed in answers 10 and 12.  This quantitative part of
the lesson confirms these conclusions: the deeper waters (locations A and B) are less sensitive to   `)(       p->*>*BB  extreme atmospheric variations than the shallow waters (C and D).  The warm air temperatures in
February 1990 result in coastal SSTs more than 3 degC higher than the climatology.  The SSTs of
deep waters are 1.5 degC higher in that month. 

In general the monthly averaged SSTs, as measured by the NOAA satellites, are a little higher than
the climatological SSTs, as measured by ships.  It is not clear whether the NOAASSTs are
systematically higher, due to the difference in measuring technique or whether the SSTs in the years
1990 and 1991 were higher than the climatology. 
In general you may conclude from this lesson that the annual variability of the SSTs in the North Sea
is high, especially in the shallow waters, and that in situations where accurate SSTs are required, the
use of climatological SSTs is not advisable.      


   
)         p->*>*BB  
# &m     P 7b&P#           p             z P    # c     P E37ApP# `"l Practical Lessons: 5p     *        p->*>*BB  
a. Practical Lessons Using Marine and Coastal Image Data


   T ` a- 5: ANALYSIS OF THE TURBID PLUME OF THE GIRONDE ESTUARYă
   T 9 7a< (FRANCE) USING SPOT DATAÃ



   T   aB Aim of Lessonă

To understand how to monitor estuarine discharge in oceans using SPOT data.



   T  aC Objectivesă

On completion of this lesson, you should be able to:

1) enhance SPOT images to study coastal features

2) identify turbidity plumes and fronts using estuarine images

   T l 3) evaluate the insitu turbidity using an arithmetic combination of radiometric values of two SPOT
channels

4) compare estuarine discharges under different hydrodynamic conditions.



   T V ~a= Background Informationă

   S  Equipment

For this lesson you require a PC running the UNESCO processing toolkit, BILKO (version 1.3). 
Some acquaintance with the use of BILKO to display satellite images is assumed.  This lesson
illustrates the application of new facilities provided by version 1.3.

Coloured pencils, squared paper and tracing paper will be useful for answering some of the questions
presented in the lesson outline.

   S " Oceanographic information

The Gironde Estuary (France) is, with the Chesapeake Bay (United States), the best studied and most
significant estuary in the world in relation to the information it has provided concerning sediment
processes and their products (Castaing and Allen, 1981; Nichols and Biggs, 1985).  The Gironde is
located on the Bay of Biscay, on the southwest coast of France (Figure 1).  The estuary itself is a
coastal plain estuary and shows a regular funnel shape with crosssections increasing exponentially
seaward from the head of tides, located 165 km from the mouth.  At its maximum, near the mouth,
the estuary is 12 km wide at high tide.  The inlet itself is characterised by two distinct channels,   _)+        p->*>*BB  separated by an extensive shoal region.  The adjacent continental shelf forms a smooth plain, gently
sloping to the shelf break (200m isobath), some 150 km off the coast.  The tides in the Bay of Biscay
are semidiurnal, and the tidal range in the inlet varies from 1.5 m (neap tides) to more than 5 m
(spring tides).

The average discharge of the Garonne and Dordogne rivers which join in the Gironde estuary is about
   T  850 m3sé1, with extreme winter and spring (November to May) flood values ranging up to 5000 m3sé1. 
   T  During summer months (June to October), the discharge decreases to less than 150 m3sé1.  The
suspended sediment flux from the rivers follows a seasonal cycle, directly related to the variations
   T  in liquid discharge.  The total annual influx of suspended silt and clay varies between 2x106 and 3x106
   T p tons.  Every year, about 1.5x106 tons are expelled to the Atlantic Ocean.  It appears that the
continental shelf does not supply finegrained sediment to the estuary.

In this lesson, we illustrate the importance of fluvial mechanisms, and their interaction with other
hydrodynamic processes, for example tides, in controlling and modulating the transport of suspended
sediment out of a macrotidal estuary.

Aerial observations and satellite imagery of the estuary mouth and the adjacent shelf have shown the
existence of large plumes of turbid water seaward of the inlet during high fluvial floods (Froidefond
   T  et al., 1991).  Here we study images from SPOT satellites.

   S  SPOT satellites

The orbital altitude of SPOT (Systeme Probatoire d'Observation de la Terre) is 830 km and the sunsynchronous orbit is similar to that of Landsat.  The southbound orbits cross latitude 40$N at 1000h
local time and ground tracks are repeated at 26 day intervals.  SPOT1 was launched by Ariane on
22nd February 1986 and SPOT2 on 22nd January 1990.  SPOT employs highresolution visible
(HRV) imaging systems which operate in either a multispectral mode (3 channels) or a highresolution
panchromatic mode.

c 
 d d x     @(   n<                   
 !d d x     R,   			           c  h 
 
H%                 S  Sensor characteristics             S  ch; Multispectral mode             S  e Panchromatic modeă  
 
` 	 	R              Angular field of view   2       hG 4$13-   2       Ek 4$13- ` 	 	8 	 	              Field of view
(Groundswath width)   j      hC 60 km x 60 km   j      g 60 km x 60 km 8 	 	8 	 	2              Detectors per spectral band
(CCD)          hG 3000          ck 6000 8 	 	` 	 	j              Ground resolution cell (pixel)   "       yhD 20 m x 20 m   "       ,h 10 m x 10 m ` 	 	` 	 	               Green spectral band   b#       QhC 0.5  0.59 m   b#       f 0.51  0.71 m ` 	 	` 	 	"              Red spectral band   $       #hB 0.61  0.68 m   $        ` 	 			b#              Near I.R. spectral band   R&       #hB 0.79  0.89 m   R&        		 $   
   T *' *a8 Table 1: Characteristics of SPOTă

   (,        p->*>*BB  Ԍ   S    Image data

Five SPOT 1 and 2 images are provided for this lesson.  The satellite data were received and
processed at the Toulouse CNES station, France.

The images have been calibrated in terms of radiometric values, geolocated and geometrically
rectified.  For this lesson, they have been resampled in our laboratory, so as to match the BILKO
specification of 512 columns by 256 lines.

The image definitions are given in Table 2.  Each image is 512 ' 256 pixels.  The filenames are
coded to help the user during the lesson: GIR for "Gironde Estuary", the following two numbers
represent the year, the third number (1 to 3) corresponds to the SPOT channel number.  Five of the
images have been obtained by selecting every sixth pixel in every twelfth row of the original data.
   T 
 GIR861P represents a section of the larger image GIR861 (Figure 1).  The P character indicates that
this image, sampling every pixel in every second row, is more precise, but of course, covers a more
restricted area (approximately 10 ' 10 km instead of 60 ' 60 km).  On the screen, all the images
are distorted in the NorthSouth direction, because 256 lines represent the same insitu distance as 512
columns in the WestEast direction.


c  
 !d d x     R,   			          
 Ad d x     -   R                    c " 				Y 
 
$"                 T   Image
   S b ! File name   :         T  D+ Satellite
   S b * Name   :         T  : Over pass
   S b 9 date   :         T  JG Channel   :         T  D[ Sampling
   S b DV (No. of Pixels)   :         T  Fn Pixel size
   S b +r (m)ă Y 
 
8 	 	                 T  GIR861.DAT          R, SPOT 1          30.4.1986          L XS1          [DY 1 out of 6 '
 1 out of 12   r      5o EW  NS
5o 120 x 240 8 	 	a 	 	:                 T  GIR862.DAT          R, SPOT 1          : 30.4.1986          L XS2                     a 	 	a 	 	r                 T [ GIR863.DAT   4       R, SPOT 1   4       : 30.4.1986   4       L XS3   4          4        a 	 	8 	 	                 T  GIR861P.DAT          R, SPOT 1          30.4.1986          L XS1          [DY 1 out of 1 '
DZ 1 out of 2   l      5o EW  NS
To 20  x  40 8 	 	h		4                 T  GIR911.DAT          R, SPOT 2          : 25.5.1991          L XS1                    5o EW  NS
5o 120 x 240 h		 l   
   T  aH Table 2: Image Description   -        p->*>*BB  Ã
   S   aA Lesson Outlineă


   T ` 1.  Enter the BILKO software, load image GIR861 and, once loaded, look at the image using Display
   T 9 image.  Use the Modify LUT function to allocate the available 16 grey levels to cover the full range
   T  of pixel values.  Do this using Histogram equalisation and display the image.  Familiarize yourself
with this image and compare it with Figure 1.  Note the different fronts and roughly evaluate the
   T  seaward extension of the turbid plume.  (Use the Cursor function in pixel steps of 20 to determine
the number of pixels, then refer to Table 2 for the pixel sizes.)

   T L	 2.  Use the toggle function through the Multiimage option to compare GIR861 with GIR862, then
   T %
 GIR861 with GIR863.  What are the main differences between the three images concerning the turbid
plume and the land?

   T  3.  Reload image GIR861, display and use the Histogram option to determine the pixel values which
   T 
 are found most frequently in the histogram.  Display the image and, using the cursor function in pixel
steps of 20, find and give the location of the corresponding areas.

   T  4.  Repeat 3, successively with images GIR862 and GIR863.  Usually, a direct relationship holds
between the reflectance at channels XS1, XS2 and the suspended matter concentrations in the water
   T  (Froidefond et al., 1991).  Which of these two channels is best suited for low concentrations (i.e.,
lower reflectances)?  What is channel XS3 useful for?

   T K 5.  Select the Image arithmetic option in the Multiimage submenu.  This option allows you to
perform simple arithmetic operations on two images.  In the case of SPOT images, it is of interest
to enhance the available information by adding channels XS1 and XS2 to study the turbid plume. 
   T  Load GIR861 (=A), then GIR862 (=B).  To add pixel values from A and B, type: a=1; b=1; c=0
   T  by pressing the <tab> key, followed by the <enter> key.  On completion, the program returns to the
   T  Main menu and the resultant image may be viewed by selecting the Display image option.

   T 6 6.  Use the Histogram equalisation option to enhance the new image.  Load EGA.PAL  by using
   T  the Load palette option and display.  Turbidity concentrations decrease from the estuary mouth
(white and yellow colours) to the open ocean (blue colours).  Several turbid fronts are visible and may
be compared with Figure 1.

   T p 7.  Use the Transect function to display two turbidity profiles across the turbid plume.  The first
should be given an EastWest orientation by moving the first cursor to 4 ' 20 lines from the bottom
left hand corner of the image and the second to 7 pixels ' 20 lines from the bottom right hand
corner, i.e. four pixels in from the left margin and seven from the right margin on the twentieth line
from the bottom.  Give a NorthSouth orientation to the second profile by moving the first cursor to
7 pixels ' 20 lines from the top left hand corner of the image and the second one to 10 pixels ' 20
lines from the bottom left hand corner.  Determine the position and steepness of the different turbidity
fronts, describe how the steepness varies along the profiles.  Where is the steepest front?  What does
it mean in terms of hydrodynamics?

   T & 8.  Using the conversion curve (Figure 3) based on insitu measurements, convert the pixel values
   T ' to suspension concentrations.  Enter the cursor function in pixel steps of twenty and make an EastWest transect across the turbid plume from the open ocean to the estuary mouth.  Convert to   (.        p->*>*BB  concentrations and plot out these values against position using graph paper, referring to Table 2 for
the pixel size.

   T  9.  Load image GIR861P and display.  Enhance the image by using the Histogram equalisation
option.  This image is less sparsely sampled than the others (refer to Table 2): it therefore covers a
smaller area, but is spatially much more precise.

Observe the swell obliquely propagating to the Medoc sandy beaches and the surf zone (white colour);
the turbid fronts and the tidal flats (brown colours) in the estuary.  On the land, features such as
roads, pine forests, fields and towns are easily recognizable.  This image has been included to
demonstrate the detail available in SPOT imagery.

   T !
 10.  Compare images GIR861 and GIR911 using the Toggle function through the Multiimage
option.  Trace the limits of the turbid fronts on the screen with tracingpaper.  Refer to Figure 2,
where the main hydrological parameters are presented (fluvial discharge and tide).  Describe the main
feature differences between the two images.  Can you say anything about the phenomena responsible
for these differences?



   T  aC Referencesă

For information on SPOT satellites:

CNES SPOT IMAGE (1986).  Manuel de l'utilisateur des donnees SPOT.  SPOT Image, Toulouse,
2 vol. (French and English)

SPOT NEWSLETTER.  Twice yearly publication.  SPOT Image, 5 rue des satellites, F31030
TOULOUSE CEDEX.

For information on availability of images:

SPOT IMAGE
5 rue des satellites
B.P. 4359
F31030 TOULOUSE CEDEX (FRANCE)

For information on the seaward dispersion of suspended sediment from the Gironde estuary:

Castaing, P. and Allen, G.P., 1981. Mechanisms controlling seaward dispersion of suspended
   T ! sediment from the Gironde: a macrotidal estuary in France. Marine Geology, 40, 101118.

Froidefond, J.M., Castaing,P., Mirmand, M. and Ruch, P., 1991. Analysis of the turbid plume of
   T M$ the Gironde (France) based on SPOT radiometric data.  Remote Sensing of Environment, 36, 149163.

   T % Nichols, M.N. and Biggs, R.B., 1985.  "Estuaries".  In Coastal Sedimentary Environments, R.A.
Davis Jr. ed.  SpringerVerlag, New York, 77186.


     a) y    b /                    d       d d                                                        /                 y    a)/         p->*>*BB  Ԍ
This lesson was prepared by P. Castaing and J.M. Froidefond, October 1992.


   R ` Comments or suggestions regarding this lesson should be sent to:

P.Castaing and J.M.Froidefond
Universite de Bordeaux I
Departement de Geologie et Oceanographie
URA CNRS 197
Avenue des Facultes
33405 TALENCE CEDEX (FRANCE)

Tel: 33  56 84 88 50 or 88 76
Fax: 33  56 840848
   0         p->*>*BB  







































   S ! ۘaH Figure 1: Location Map   !1         p->*>*BB  






































   T   aH Figure 2: Main Hydrological Parameters    2         p->*>*BB  



































   T ` aH Figure 3: Curve for Conversion of Pixel Values to Suspended Sediment Concentration   `3         p->*>*BB  
   T   a> Answers to Lesson 5 

1.  The turbid waters spread out in patches limited by fronts around the mouth of the Gironde.  The
extension of the turbid plume is approximately 45 kms EastWest and NorthSouth.  These distances
can also be measured in Figure 1.

   T  2.  The GIR861 image shows the maximum extension of the turbid plume.  The coastal pine forests
   T  are evident.  In the GIR862 image, the maximum extension of the turbid waters westwards is no
longer visible.  On the other hand, several levels of turbidity can now be seen in the estuary.  In the
   T s same way, the sandy beaches appear more clearly.  In the GIR863 image, the turbid plume is no
longer visible.

   T 
 3.  The histogram of the GIR861 image shows three frequency peaks.  The largest peak (N=4548)
corresponds to the moderately turbid zones, situated at the fringe of the turbid plume.  The second
peak (N=5358) corresponds to the most turbid waters close to the mouth and in the Gironde estuary. 
The third peak (N=3640) corresponds to the slightly turbid waters to the West (for the least turbid
waters completely westward, N=3132).  It is to be noted that on land, N=3640 corresponds to the
   T 5 pine forest.  The histogram of the GIR862 image shows only one frequency peak (N=2022) which
corresponds to the moderately turbid waters that are not visible on the screen in this channel, at the
   T  fringe of the turbid plume.  The histogram of the GIR863 image shows a very sharp peak (N=89)
which corresponds to the moderately and slightly turbid waters (N=13 for the most turbid water of
the estuary).  The second very flattened maximum of the histogram (around the maximum N=58)
corresponds to the emerged zones.

4.  It is the XS1 channel that is best adapted to give the maximum amount of information on those
waters where the concentrations of solid matters are low.  In this channel, the reflectances, for this
type of water, are higher than in the XS2 channel.  On the other hand, the XS2 channel provides
more precise information than the XS1 channel for water in which the turbidity is high.  The XS3
channel is extremely useful for finding the precise limit between the water and the land, as the
wavelengths of this channel (0.790.89 m) are massively absorbed by the water which then shows
up dark.

7.  On the EastWest profile, the pixel values (ie. the concentrations in suspended material)
progressively increase towards the estuary.  A front limit is clearly marked in the West between the
clearer and slightly turbid waters.  Two other less distinct fronts are visible eastward within the turbid
plume.

These various fronts indicate several phases of discharge of suspended material over the 48 hours
preceding April 30, 1986.  At each ebb tide, silts and clays are driven seawards in the form of a
turbid plume.  From one ebb tide to the next 12h 25min later part of the discharged sediment remains
in suspension and is transported seawards under the action of currents.  The turbid plumes discharged
on the following ebb tides are partly superimposed over the previous ones, while at the same time
   T O$ gradually pushing them several tens of kilometres away from the mouth.  In the GIR861 image, at
least three discharge phases can be recognized, the oldest being roughly 40 kms from the mouth.

The NorthSouth profile cuts through part of the turbid plume perpendicularly.  The concentrations
decrease progressively and regularly towards the South.  There is a sharp front in the North (it is the
steepest) indicating a very distinct limit between the turbid waters of the turbid plume and the "clear"
waters of the continental shelf.  This sharp limit indicates that, when the image was recorded, the   `)4        p->*>*BB  currents were running in a SENW direction off the mouth of the estuary.  At this phase in the tide,
   T   the turbid waters move in a northwesterly direction.  This was confirmed by insitu current
measurements.

10.  In the two images, the outer front of the turbid plume is marked by a strip in red.  In the
   T : GIR911 image (28051991) the turbid plume has not left the estuary and the turbidities on the
   T  continental shelf are low (check by an EastWest profile using the Transect option).  This image is
   T  thus totally different from the GIR861 image (30041986) where the outer turbid front is 45 kms off
the mouth of the estuary.  The explanation for this is given in Figure 2.  In both cases, the tide height
   T  is near the mean (b), but the GIR861 image was recorded at midebb, whereas the GIR911 was
recorded at the beginning of the flood.  What is even more indicative is that the fluvial flow rates (a)
   T N	 are very different.  GIR911 corresponds to a period of average flow rates (900 m3sé1) and there had
been no high flood in the previous months.  On the other hand, GIR861 corresponds to a period of
   T 
 relatively high fluvial flood (>2000 m3sé2) and the estuarine waters are extremely turbid.  At each
ebb, part of the suspended material is discharged into the ocean, thus increasing the mean turbidity
of the water of the continental shelf.
   
5         p->*>*BB   # &m     P 7b&P#           p       /     z P    # c     P E37ApP# `"l Practical Lessons: 6p      6        p->*>*BB  
a. Practical Lessons Using Marine and Coastal Image Data


   T ` a+ 6: DETECTION OF SEA SURFACE LIFE WITH AN AIRBORNE SYNTHETIC
aA APERTURE RADAR.



   T   aB Aim of Lessonă

To gain familiarity with digital images of fishing events as recorded by an airborne Synthetic Aperture
Radar (SAR).



   T 
 aC Objectivesă

On completion of this lesson, you should be able to:

1) display SAR image data so as to reveal fishing events such as tuna schools or a tuna purse seiner
during its fishing operation and live events, such as the movement of whales

2) interpret the texture of a SAR image.



   T  ~a= Background Informationă


   S T Equipment

For this lesson you require a PC running the UNESCO image processing toolkit, BILKO (Version
1.3).  The most effective display of SAR images is achieved by running BILKO with a VGA display,
although the images used in this lesson can still be viewed, but less effectively, using an EGA display.


   S < Information about sea surface life and tuna fishery 

Tuna schools, from time to time, act in such a way as to change the appearance of the sea surface. 
(The terms boiler, breezer, jumper... are sometimes used).  The way they do this depends on the
species and the school size, but the behaviour is well known and may be detected from a boat,
helicopter or aircraft.  Fishing boats and nets, as well as the tuna schools moving in surface or
subsurface waters, generate surface distortions which should be detectable by a SAR radar system.


   T & HAREM experiment 

The image datasets supplied with this lesson are all derived from the HAlieutics Radar
Experimentation Mediterranean (HAREM), an experiment conducted from 1st to 14th August 1989   ])7        p->*>*BB  in the Golfe du Lion off the western Mediterranean coast of Europe (south of France) with the aim
of testing the potential of SAR imaging theories to applications in halieutics.  The SAR data were
provided by the aerial E-SAR (high resolution radar device from the German aerospace agency, DLR)
flown on a Dornier 228 aircraft flying at 2000m. 

Concurrent ground-truth data of tuna school size, cetacean identification, and fishing device
positioning are necessary to allow a rigorous examination of SAR signatures.  This calls for the
synchronous information from the tuna purse seiner fleet and its aerial survey plane.  It has proved
possible to identify some targets clearly: boats, nets and, above all, cetaceans and tuna schools.  The
weather conditions were good.

HAREM experiment data suggest a new line of research for fisheries management and the survey of
marine surface life, in particular for cetaceans and pelagic fish.  Airborne SAR is potentially better
than any other monitoring technique since it covers a large area, is not influenced by clouds and is
independent of the fishing method in use.  The aim is not to develop a new tool to assess the catch
but one to assess the tuna potential of an important area, fished or unfished, and to provide
information for fisheries management through surface stock assessment.


   S  Synthetic Aperture Radar

The theory of the SAR imaging of the sea surface is briefly presented by Robinson in the TREDMAR
Marine Image Data Module 2 (Practical lesson 1, SEASAT-SAR images of the "Pas de Calais" -
France).  The HAREM data include, for the first time, C-band SAR images of the ocean surface
showing fishery activity, tuna schools and the presence of marine mammals.  The Cband radar
operates at 5.3GHz with VV (vertical transmit and receive) polarization.  The wavelengths are 5.6cm
for C-band and 3.2cm for X-band.  These wavelengths are important because the new European
satellite ERS-1 is equipped with a C-Band SAR.

Analysis of the HAREM data indicates that marine surface life such as tuna schools (bluefin tuna -
   T P Thunnus thynnus), marine mammals and fishing activities (purse seiners) generate SAR signal
modulations of comparable, and detectable, magnitude in the Cband.


   S  Image Data

Three image datasets are provided for this lesson.  These images were digitally correlated
(reconstructed from the raw data coming from the airborne radar) and supplied by the SAR processing
facility at CNES, Toulouse, France.  The processing has produced a geometrically undistorted image
with pixels which correspond to squares on the ground of 0.7 m or 1.0 m.

The image files are denoted SCHOOL.DAT, SEINE.DAT and WHALES.DAT.  Each comprises 256
' 256 pixels, SCHOOL.DAT and SEINE.DAT having a pixel size of 0.7 ' 0.7 m and
WHALES.DAT a pixel size of 1.0 ' 1.0 m.





   Z)8         p->*>*BB  Ԍ
   T   aA Lesson Outlineă


   T a 1. Display image SCHOOL.  First use the Load image function, noting that in BILKO Version 1.3
this presents the user with a sub-menu from which to select the part of the original image file to be
displayed.  In this case, every pixel should be displayed. Once loaded, look at the image by using
   T  Display image.


   T s 2. Use the Histogram function to examine the distribution of the pixels, you should find that it is
Gaussian.  This distribution (Figure 1) is typical of SAR images (See the practical lesson 1 in the
Marine Image Data Module 2; SEASAT-SAR images of the "Pas de Calais" - France).  In the upper
   T 
 left of the image a tuna school (bluefin tuna) may be seen.  Use the Cursor function to determine
typical values for the school.  Note the values of the thirteen pixels between A (30,26) and B(30,38). 
The higher values are caused by the fishes in the school. The radial gradient of the backscattered
radiation is due to the higher density of jumping fishes in the middle of the school.  Examine the
value of pixel (30,37): what is special about this value?


   T  3. Select the school in the upper left of the image using the function Subimage in Modify image or
   T  load the image SCHOOL again to select only the tuna school by setting First line to 20 and Last line
   T  to 60, the First pixel to 1 and the Last pixel to 60.  Use the Histogram function to examine the
distribution of the pixels corresponding to the tuna school (Figure 2). The distribution in the
histogram is not typical of SAR images.  This histogram exemplifies the dispersion of the pixel values
representing the school.


   T  4. Display image SCHOOL and experiment to show the difference between the colour and
   T  monochrome grey scale displays obtained by selecting the EGA.PAL or the VGA.PAL with the Load
   T \ palette function.  For SAR images, the monochrome grey scale palette VGA.PAL is the best display
for interpretation.  If the school is correctly identified with the colour palette (EGA or TEMP1), on
the other hand, the variation of the roughness of the sea surface (on the right of the school)
disappears.


   T m 5. Load SEINE using the Load image function.  This image represents a purse seiner with its net. 
In the small scale tuna fishery in the western part of the Mediterranean Sea, boats are 22 to 25 meters
long with nets of about 1800 meters.  In this image, the tuna boat is the clear spot in the lower left
part of the image.  The net looks like a circle.  Inside the net can be distinguished two white points:
the tuna school and the "speed boat", a small boat which assists the purse seiner in its fishing
operations.  What additional features can be detected on this image?  To pick out the net, first use
   T ~# the Histogram equalization in Modify LUT menu and then use the Smooth filter. 


   T & 6. Load WHALES using the Load image function.  In the image, two whales may be seen swimming
   T & side by side.  Use the Cursor function to estimate the lengths of the whales.  Note the values of the
pixels representing the whales and compare them with those in the tuna school pixels.  Select the
   T ( Gradient option in the Filters function.  This filter emphasises the contours of the whales.   (9        p->*>*BB  Ԍ

   T  aC Referencesă


For information on the processing of SAR data. 

   T  Horn, R., 1989. Cband SAR results obtained by an experimental airborne SAR sensor. Proc.
   T  IGARSS'89 Symposium, Vancouver, Canada, July 1989, IGARSS edition, 566568.

   T u Vaillant, D. 1984. Traitement des donn)es du radar ! synth/se d'ouverture, in Cours de
   T O	 math)matiques spatiales, mars 1984, CNES (Toulouse, France), CEPADUES Editions (Toulouse,
France), 9831013.


For information on HAREM experiment

Petit, M., J.-M. Stretta, H. Farrugio et A. Wadsworth, 1990. HAREM. Halieutique et Radar,
Exp)rimentation en M)diterran)e. Potentialit)s du radar SAR en halieutique. Application ! la p+che
   T  thoni/re de surface et ! la p+che artisanale. Collection Etudes et Th/ses, Editions de l'ORSTOM.
Paris 1990. 122p.


     s y    b                     d       d d                                                        :                 y 


This lesson was prepared by J.-M. Stretta, M. Petit, H. Farrugio and A. Wadsworth

   R  Comments or suggestions relating to this lesson should be sent to:

Dr. J.-M. Stretta
Centre ORSTOM-UF Hydrobiologie-Oc)anographie
B.P. 5045-34032 Montpellier c)dex 1, FRANCE
e-mail: stretta@orstom.orstom.fr
Tel: 3367617400
Fax: 3367547800

Dr. M. Petit
Centre ORSTOM-Surveillance de l'Environnement Assist)e par Satellite. 
BP 172-97492 Sainte Clotilde, FRANCE
e-mail: petit@orstom.orstom.fr

Dr. H. Farrugio
Centre IFREMER-1 rue Jean Vilar-34200 S/te, FRANCE

A. Wadsworth
GDTA 18 Avenue E. Belin-31055 Toulouse, FRANCE


   (:         p->*>*BB  Ԍ














   T  aD Figure 1:


















   S  aH Figure 2:   ;         p->*>*BB  
   T   a> Answers for Lesson 6 ă


2. The difference between a sea pixel and a school pixel is about 200. The values of the thirteen
pixels between positions A and B are:
57, 73, 100, 105, 186, 131, 162, 146, 126, 237, 209, 255, 211.

The value of the pixel (30,37) is 255. This maximum value corresponds to a saturation of the signal.

5. The swell is visible in the image. 

6. Some pixels showing the whales also show a saturation of the signal. This is probably caused by
reflection from water on the skin of the whales.
    <         p->*>*BB    /       p       wU   D'  z P    # c     P E37ApP# `"l Practical Lessons: 7p # &m     P 7b&P# 
a. Practical Lessons Using Marine and Coastal Image Data


   S ` +a- 7: BATHYMETRY PREDICTION FROM SATELLITE ALTIMETER DATA



   T   aB Aim of Lessonă

To predict bathymetry from gridded radar altimetry measured by earthorbiting spacecraft over remote
oceans where conventional ship borne depth soundings are sparse but satellite altimetric measurements
are dense and evenly spaced.



   S 
 aC Objectives

   T 1 On completion of this lesson you should be able to:

1) distinguish between the surface measured by radar altimetry, the geoid and the gravity field

2) predict bathymetry from radar altimeter data

3) identify from altimeter data tectonic features which are not visible on archival bathymetry data
bases

4) decribe the relationship between short wavelength gravity field anomalies and bathymetric features
in the oceans

5) compare the predicted bathymetry with the archival bathymetry data set

6) gauge how much further improvement can be made when bathymetry predicted from satellite
altimetry is combined with the archival data set.



~a= Background Information

   S   Radar Altimetry

The radar altimeter mounted on earthorbiting satellites emits pulses which are reflected from the sea
surface.   Monitoring the precise position of satellites via ground tracking allows the distance between
the altimeter and the instantaneous sea surface to be determined.  The sea level surface is modified
by density inhomogeneities within the earth, including sea floor topography.  Measurements of the
sea level topography, however, must be corrected for oceanographic "noise", such as currents, tides,
and atmospheric pressure variations, which also warp the ocean surface.

Upon correction for these oceanographic effects, the residual equipotential surface is termed the
"geoid".  The geoid undulates about the reference ellipsoid in a way which may be described by a   Y)=        p->*>*BB  series of spherical harmonics of specified wavelength and amplitude.  A large seamount may cause
a sea level high as much as 50 km across and 1 metre in amplitude.

The most recent altimeter working at the time of writing this lesson is the ERS1.  The US Navy's
Geosat was operative from March 1985 to January 1990.  Earlier altimeters were the shortlived
Seasat between June 1978 and October 1978, and Geos3 between April 1975 and July 1978.  Geosat
was launched on March 12, 1985 into an 800 km altitude, 108 inclination orbit that generated 3day
nearrepeat ground tracks.  The Geodetic Mission (GM) generated a dense data set, with subsatellite
tracks of the order 24 km apart.  At the completion of the 18monthlong primary mission, it then
changed its mission to a  17day Exact Repeat Mission until its failure on January, 1990.  Its footprint
diameter was 4 km.  In this lesson, highpass filtered Geosat data are used.  Wavelengths exceeding
about 500 km have been removed, and wavelengths of 116 km reduced by 50 % in amplitude.
 
U           1                                           1                                 
r!  #bh
:                 B  d d `                                                    >                } K    # i x    P 7P# 
 Relationship between the geoid and the gravity field

The geoid undulates about the reference ellipsoid due to density inhomogeneities within the
earth.  Gravity anomalies also occur due to the same inhomogeneities which cause the geoid
undulation.  However, their responses are different: the geoid undulations are proportional to
the inverse of the distance between the reference ellipsoid and the mass anomaly, whereas
the gravity anomalies are proportional to the inverse square of the distance.  Consequently,
the geoid undulations show broader warping than the gravity anomaly curve over a point
mass anomaly at depth.  Over a region of mass anomaly, there is an additional potential U,
and the relationship with the geoid undulation is written as:


2 gN= U

where N is the warping of the geoid from the reference ellipsoid, and g is the mean gravity. 
The gravity anomaly is the normal component of the gradient of anomalous potential U.r$  ""
""X! "$ 

   S   a< Data used in this lesson

   T  There are two primary sources of data for this lesson, ship based observations of ocean depth and
measurements of the ocean geoid derived from satellite altimetry.

Three basic image datasets are provided along with several others derived from them:

(i) The archival bathymetry data were derived from DBDB5, a global digitally gridded bathymetry
   T ! data base of shipbased observations.  This dataset is supplied as a basic data image (DB5TOP) and
   T " as lowpass (LPTOP) and highpass (HPTOP) filtered images.

   T J$ (ii) Highpass filtered GEOSAT GM data (GEOSAT) provided by the US Naval Oceanographic
Office .  This represents geoid anomalies relative to the reference ellipsoid.

Both data sets (i) and (ii) were supplied gridded at 5- by 5- intervals and have been regridded to a
common 10km by 10km cell size using a minimum curvature technique.

   T [) (iii) Bathymetry predicted from the satellitederived geoid (PRDTOP), using physical principles 0  [)>         p->*>*BB :"i  ! > 0  described in the text boxes.  Since bathymetric (topographic) and geoid anomalies are well correlated
in the wavelength range from about 15 to 200 km, we first predicted highpass bathymetry from the
Geosat geoid anomalies only and added them to the regional bathymetry (represented by the lowpass
DBDB5 in this study).  We used a simple no compensation model for this lesson.

The other image files are worked examples resulting from applying Bilko functions to (i), (ii) and
(iii).
	A  #b!*                 <   d d N  FILE2                                             ?   %.        # i x    P 7P# 
  } K   
 Relationship between bathymetry and gravityă

Gravity anomalies are closely correlated with bathymetry in the wavelength range from about
  } K   15 to 200 km (Talwani et al., 1972; McKenzie and Bowin, 1976; Watts, 1978; Vogt and
Jung, 1991; Jung and Vogt, 1992).  Their relationship is best studied in the wave number
domain.  Assuming the relationship is linear, it can be expressed as:

z
- G(k) = Z(k)B(k)

where G, Z, and B are the discrete Fourier transforms of gravity, transfer function, and
bathymetry, respectively, and k is the wave number defined as k=2p/l, where l is the
anomaly wavelength.

The transfer function, Z, thus can be obtained by dividing G by B.  In the presence of noise,
the transfer function is better estimated by dividing the crossspectra of the gravity and
bathymetry by the power spectrum of bathymetry (McKenzie and Bowin, 1976):

n	
& Z(k) = G(k) B*(k)/ B(k) B*(k)

where * denotes the complex conjugate.

However, we usually do not have enough gravity and/or bathymetry measurements to derive
the transfer function in remote oceans.  In the case of this lesson, we do not have enough
bathymetry data, while good geoid coverage is available from Geosat altimeter data.
(Shipborne bathymetric profiles are widely and randomly spaced at about 100200 km in the
southern Pacific Ocean whereas Geosat altimeter profiles are closely and evenly spaced at
about 24 km.) The geoid anomaly  associated with a gravity anomaly can be expressed as:


- H(k) = G(k)/gk
 
where H(k) is the geoid anomaly in the wave number domain, and g is the average gravity. 
Hence, we use a theoretical model transfer function derived under certain assumptions. 
Among many theoretical models, the following three models were considered for our study;
1) no compensation which assumes that oceanic crust is strong enough to support seamount
loads and thus is not deformed by the loads, 
2) Airy type of local compensation, which assumes that elevated seamounts are underlain by
crustal roots of less than normal density, and 
3) platetype regional compensation, which assumes that the "mechanical lithosphere", of
which the crust is the uppermost part, responds as an elastic plate capable of supporting
  } K @ shear stresses produced by seamount loads.  The first was used to derive PRDTOP.	$  """"A "$ Ԓ   S    Image data

The images used in this lesson are listed in Table 1.  All images have dimensions of 245 lines ' 313
pixels, each pixel corresponding approximately to a 10 km ' 10 km square.  The ocean depth ranges
from 250 meters to 6045 meters and highpass geoid anomalies range from 80 cm to +117 cm in
this lesson area.  The image area (130$W  180$W, 45$S  65$S) is shown in Figure 1.  Palette files
   T  TOPO.PAL and GEOID.PAL are also supplied.

   T  

a4 Table 1: Images supplied for this lesson.ă

^  
 Ad d x     -   R                  
 aD	d d x     r@   M        ^  h	
l                 T 	 
% Filename   #
          T 	 wH Type Descriptionă 
` 	r              '
( DB5TOP   R       wC Gridded Archival Bathymetry ` 	` 		              6
( LP_TOP          wB LowPass Archival Bathymetry ` 	` 	R              ,
( HP_TOP          wB HighPass Archival Bathymetry ` 	` 	              
( GEOSAT   r       Dw> HighPass Geosat (GM) Geoid Anomalies ` 	` 	              
( PRDTOP          IwF Predicted Bathymetry ` 	` 	r              4
( DIFPRD   2       w> Differences between DB5TOP and PRDTOP ` 	` 	              S
) SMDB5          wB A smoothed version of DB5TOP ` 	` 	2              
( GRADB5          w> Highpass filtered version of DB5TOP ` 		              >
( DIFDB5          0w? Difference between DB5TOP and SMDB5 		    



   T  aA Lesson Outlineă

A Examining the Archival Bathymetry

   T C 1. Load the archival bathymetry image data (DB5TOP) and display it.  Load the provided palette
   T  (TOPO.PAL) or create a suitable palette  to show the regional variability of bathymetry.

Look at the midocean ridge.
Does it have a central rift valley?  Where?
What is the trend of the midocean ridge in this area?
Where are the ocean basins found?
How are they different from the midocean ridge area?
Which area shows a rougher ocean floor topography?  Why?
Locate the fracture zones.
What is the general trend of the fracture zones in this area?
How is their trend different from that of the midocean ridge?
How are the fracture zones formed?
Where are the seamounts found?
   )@        p->*>*BB  Ԍ   T    2. Using the Transect function, display a few profiles across the midocean ridge and the transform
   T   faults.  Using the Cursor function, look at the image values of the seamounts. 

Keep in mind that depth is positive downward from the sea surface so profiles are seen as mirror
images.  Thus, the shallower the depth, the lower the pixel image value.  You can make a normal
   T : profile by using the Multi image, Image Arithmetic option.  First, load DB5TOP and load any
   T  dummy file, e.g. SMDB5.  Then set A=1, B=0, C=255, and use the Transect menu.

   T  3. Enter Histogram and find the mean value of DB5TOP.  Then, using the information that the
depths of maximum (6045 meters) and minimum value (250 meters) have been mapped onto the
interval from zero to 256 integers, estimate the mean depth of the lesson area.

   T %
 4. Now apply the Filters, Smooth option over and over (about 5 times or more) to the archival
   T 
 bathymetry (DB5TOP) and save the smoothed image (a smoothed image file SMDB5 is provided)
   T  to the hard disk or another floppy disk.  Alternatively enter the Filters, Gradient option and save the
   T  highpass filtered image (a highpass filtered image GRADB5 is provided) to the hard disk or another
   T 
 floppy disk.  Repeat the Transect and the Cursor commands across the midocean ridge, the fracture
zones and the seamounts on the lowpass and highpass filtered images.

   T  5. Enter the Multi image, Image arithmetic option with  DB5TOP and SMDB5 (or your own saved
   T  image in stage 4).  Set  A=1, B=1, and C=0.  Save the difference image (a  difference image file
   T  DIFDB5 is provided).

   T t 6. Enter the Multi image, Toggle option in order to compare the provided lowpass and highpass
   T M archival bathymetry images (LP_TOP and HP_TOP) with the saved images, SMDB5 and DIFDB5
   T & (or GRADB5) respectively.  Are the differences between them significant ? 

   T  7. Enter the Multi image, Toggle option between  DB5TOP and LP_TOP (or SMDB5 from stage
4).  What are the smoothing effects, particularly, on the fracture zones ?


B.  Examining the altimeterderived geoid

   T  8. Load the highpass Geosat satellite altimeter derived geoid anomaly image (GEOSAT) and display
   T  it.  Load the provided palette (GEOID.PAL) or create a suitable palette to highlight the fracture
   T  zones and the seamounts.  You can apply the Modify LUT (either the Linear stretch or the
   T s Histogram equalization) if you wish.

Look for the fracture zones and the seamounts.
Where are they found?
Is the general trend of the midocean ridge?
   T " Why is it different from that of DB5TOP?

   T ]$ 9. Enter the Multi image, Toggle option to compare the Geosat altimeter derived geoid anomaly
   T 6% image (GEOSAT) with the archival bathymetry image derived from DBDB5 (DB5TOP).

   T & 10. Repeat stage 9 for the pair of GEOSAT and HP_TOP images.
Which shows more detailed fracture zones and seamounts?
   (A         p->*>*BB  ԌC.  Examining the bathymetry predicted from the satellitederived geoid

   T  11. Now Load the predicted bathymetry (PRDTOP) and display it.  Load the palette provided
   T  (TOPO.PAL) or create a suitable one.  PRDTOP was obtained by adding the predicted bathymetry
   T b from highpass filtered satellite altimeter data (GEOSAT) to the regional bathymetry.  Which image
file represents the regional bathymetry in this lesson?  Repeat the questions of stage 1 for this image
   T  PRDTOP.

   T  12.  Enter the Multi image, Toggle option to compare the archival DBDB5 bathymetry (DB5TOP)
   T  with the predicted bathymetry from Geosat satellite altimeter data (PRDTOP).  What are the major
improvements you observe?
 
   T &
 13.  Enter the Multi image, Image arithmetic option with DB5TOP and PRDTOP.  Set  A=1, B=
   T 
 ԩ1, and C=0.  Alternatively a difference image file DIFPRD is provided.  Display the difference
   T  image with the Modify LUT (either the Linear stretch or the Histogram equalization).  Suggest
possible causes of these differences?   


   S 9 aC Referencesă

Jung, W. and P. R. Vogt, 1992.  Predicting Bathymetry from GeosatERM and Shipborne Profiles
   T  in the South Atlantic Ocean.  Tectonophysics, 210: 235253.

McKenzie, D. and C. Bowin, 1976.  The Relationship between Bathymetry and Gravity in the
   T K Atlantic Ocean.  J. Geophys. Res., 81: 19031915.

Talwani, M., Poppe, H. and P. D. Rabinowitz, 1972.  Gravimetrically Determined Geoid in the
   T  Western North Atlantic. In: NOAA Tech Rep. ERL228AOML 72, part 2, NOAA, Boulder, Colo., 
pp 134.

   T _ Vogt, P. R. and W. Jung, 1991.  Satellite Radar Altimetry Aids Seafloor Mapping.  EOS Trans.
   T 9 AGU, 72: 465,468469.

Watts, A. B., 1979.  An Analysis of Isostacy in the World's Oceans. 1. HawaiianEmperor Seamount
   T  Chain.  J. Geophys. Res., 83: 59896004.

   T u The Ocean Basins: Their Structure and Evolution by  The Open University published  in association
with Pergamon Press contains a topic on satellite bathymetry ( See pages  1821, 4349).


     ! y    b (                    d       d d                                                        B                 y 


This lesson was prepared by W. Y. Jung and P. R. Vogt, September 1992.
   o$B         p->*>*BB     R    Comments or further suggestions should be sent to:

Dr WooYeol Jung
Code 7420, Naval Research Laboratory,
Washington, D.C. 203754 hh= F O ppX 
U.S.A.

Fax: (1) 2027670167






































   S & Figure 1: `	`	# A World Map of major plates and midocean ridge system.  The numbers are
spreading rates in mm/year.  Lesson area includes part of the PacificAntartic
   S ( Ridge (dotted small box)   (C        p->*>*BB     S    "`	a> Answers for Lesson 7ă

1. The midocean ridge system is a continuous, almost wholly submarine mountain range branching
throughout the world oceans.  It is a broad, fractured swell, in some areas smooth, in others
exhibiting a central rift valley and extremely rugged topography.  The midocean ridge typically rises
23 km above adjacent basins and is typically 10002000 km in width.  The ridge is continuously
maintained by the moving apart of lithospheric plates.  In the lesson area the Pacific plate is moving
northwest (parallel to the fracture zones) away from the Antarctic plate.  New oceanic lithosphere is
formed at the plate boundary.  This new lithosphere is hot and thermally expanded, which explains
why the midocean ridge is a "ridge."

Part of the midocean ridge system is shown in this lesson area as reddish to yellowish colours (based
   T  
 on TOPO.PAL), trending in a SWNE direction.  Water depths become gradually deeper away from
the midocean ridge system towards the adjacent ocean basins.  The ocean basins exhibit smoother
bottom topography, in part due to sediments which tend to "pond" in topographic lows, and so
diminish relief.

The fracture zones are elongate zones of unusually irregular topography that often separate regions
of different character.  Such fracture zones commonly cross, and apparently displace the midocean
ridge by faulting.  They are formed by different rates of sea floor spreading between adjacent crustal
blocks and show the past spreading directions.  Their trends are generally perpendicular to those of
the midocean ridge.  One major fracture zone named the Eltanin Fracture Zone (see Figure 1 and
this image) offsets the midocean ridge in the central part of the lesson area and trends NWSE. 
Several seamounts are also found in the lesson area, where they show as small isolated reddish or
yellowish patches.

A median rift valley is either absent or, where present, is poorly developed (thin NE trending yellow
line in the middle and northeastern parts of the midocean ridge).  This would be expected at
moderately high rates of plate separation ( about 8 cm/yr).  At slower rates the rift valley would be
deeper and the geoid low more prominent.  You can check the existence of rift valley in the lesson
   T Q area by using the Transect menu across the ridge.

3. All depths between the maximum and minimum values are mapped to the 8bit unsigned integers
from zero to 255.  Thus by multiplying 22.6 [=(max depthmin depth)/256] by the mean value of
the histogram (172.5776), you can estimate the mean water depth in the lesson area, as being about
3900 meters.

6. Some differences may occur due to the different cutoff wavelengths applied during the filtering
processes, but they should show at least some corresponding highlights.

7. Smoothing of the topography makes the fracture zones less traceable farther away from the midocean ridge axis.  

8. The fracture zones and seamounts are now more clearly seen on this image, but it does not show
   T "% the broad regional midocean ridge area because of highpass filtering on the GEOSAT geoid image.

   T & 10. The GEOSAT image shows more detailed features.

   T ( 11. The lowpass filtered archival bathymetry image LP_TOP shows the regional bathymetry.
   ])D         p->*>*BB  Ԍ12. Some fracture zones which are not visible on the archival bathymetry are clearly registered on
the predicted bathymetry.

13. The differences are mainly due to the fact that the archival bathymetry data base DBDB5 used
in this lesson was generated from widely spaced shipborne tracks while the predicted bathymetry was
based upon dense, evenly spaced satellite altimeter data.
