Storm-Relative Data Assimilation

Principal Investigator:

• Altug Aksoy (HRD/- CIMAS)

Project Members:

• Sim D. Aberson (HRD)
• Bachir Annane (HRD/- CIMAS)
• Robert G. Nystrom (HRD/NOAA Hollings Scholar)
• Kathryn J. Sellwood (HRD/- CIMAS)

The goal of this project is to make the best use of tropical cyclone observations in a storm-relative framework. The motivation comes from the possibility that a storm can move by a significant distance on the time scales of data assimilation. The figure below illustrates this nicely:

Using the Atlantic hurricanes' climatological storm speed population in panel (a), the frequency distribution of all Atlantic hurricanes is plotted in panel (b) as a function of normalized 2-h (typical half width of observation collection window in hurricane flights) distance traveled, where normalization is by the radius of maximum wind. This reveals that for about 40% of all Atlantic hurricane cases, storms move by a distance of more than one RMW, a situation hypthesized to result in large and unrealistic cross-gradient innovations in data assimilation. The storm-relative observation processing discussed in Aksoy 2013 is proposed to minimize the detrimental impacts of storm motion in such cases.

• Since the 2012 hurricane season, the storm-relative observation processing capability has become standard in HRD's HEDAS. The details of this technique are discussed in Aksoy 2013.

• Aksoy, A., 2013: Storm-Relative Observations in Tropical Cyclone Data Assimilation with an Ensemble Kalman Filter. Monthly Weather Review, 141, 506-522.