Paper on how data from small unscrewed aircraft systems can improve tropical cyclone forecasts published in Weather and Forecasting - NOAA's Atlantic Oceanographic and Meteorological Laboratory

This study represents the first time that observations from a small uncrewed aircraft system (sUAS) have been assimilated into the Hurricane Weather Research and Forecast model (HWRF) used by specialists at the National Hurricane Center (what we call an operational model). Including these data has potential for improving initial analyses and forecasts of tropical cyclones. The data were obtained using the Coyote sUAS, but these results are expected to be applicable to newer platforms which will be operational soon.

Observations without Fear: NOAA's Drones for Hurricane Hunting – The Front Page

Joe Cione of HRD holding a Coyote sUAS. A NOAA P-3 Hurricane Hunter aircraft is in the background.

Tropical cyclone analyses and forecasts generally suffer from a lack of data, especially near the surface where Hurricane Hunter aircraft cannot fly. This study looks at the potential of observations obtained using the Coyote sUAS that were deployed from a NOAA P-3 Hurricane Hunter aircraft in 2017 Hurricane Maria for improving tropical cyclone forecasts and analyses. The goal is to demonstrate the benefits of these types of observations and to provide guidance for the use of data from newer sUASs which are expected to be available soon. This is the first study that examines the data in multiple cases as they would be used in operations.

Figure 1: GOES 13 visible satellite images of the core of Hurricane Maria overlain with the Coyote wind-speed observations used in this study. The ten-digit number represents the date and time (year, month, day, and hour in UTC) of the image. C1 through C6 represent the six Coyote flights.

Three sets of model runs were conducted, one where all available data except for those from the Coyote were assimilated and two which also assimilated Coyote data. Since the Coyote flies slower than crewed Hurricane Hunter aircraft, it can gather data closer together in space. One set of runs was made with Coyote data that were thinned to a similar spacing and using the same quality control (QC, a process in which possibly erroneous data are removed) as conventional aircraft observations. Another set of runs was made to use as much Coyote data as possible, by increasing the horizontal resolution, and relaxing the QC for wind observations. The resulting analyses and forecasts were then compared with the National Hurricane Center’s best track, which gives a best estimate of Maria’s size, track, and intensity over the hurricane’s life span.

Figure 2: Average forecast errors from HWRF with no Coyote data (H221 – black), Coyote data used like other Hurricane Hunter observations (H2C1 – red) and most Coyote data (H2C2 – green) for track (left), maximum sustained wind speed (middle) and minimum pressure (right). Note that the green and red lines are usually below the black line in the graph, showing forecast improvements due to the use of the Coyote data. The forecast times extend from 0 h (initial condition) to 120 h (5 days) into the future. The numbers of cases are in blue.

■ Important Conclusions:

1) The addition of sUAS data improves forecasts and analyses of tropical cyclones using an operational model.

2) Analyses which include sUAS data are more consistent with other reconnaissance observations than those that do not include sUAS.

3) Including sUAS data reduces the tendency of the model to produce a storm which is too weak.

4) The sUAS observations can be assimilated at higher resolution with more relaxed quality control than conventional aircraft observations.

5) sUAS flight patterns should ideally be symmetric about the tropical cyclone center for the best results within the current operational system.

For more information, contact aoml.communications@noaa.gov. The study can be found at https://doi.org/10.1175/WAF-D-22-0214.1. Partial funding support was provided through the Cooperative Agreement NA67RJ0149 between NOAA and the University of Miami.