The Impact of GOES Derived Motion Winds on Tropical Cyclone Intensity Forecasts

Methods to estimate atmospheric wind speed and direction using cloud- and water-vapor-drifts observed in satellite images (Derived Motion Winds, DMWs) were developed shortly after the first geostationary weather satellite data became available in the 1970s. The DMWs from the GOES-R series satellites take advantage of the improved horizontal, temporal and spectral resolution of these current generation GOES systems and produce up to five times more data than the previous generation GOES DMWs.

In this study, operational DMWs from GOES-16, -17 and -18 are being used to calculate several diagnostic parameters related to tropical cyclone (TC) intensity changes. The hourly DMWs are composited in storm-relative coordinates over three pressure layers (outflow layer 150-300 hPa; steering layer 300-850 hPa, lower-layer 850-1000 hPa) and then objectively analyzed on an evenly spaced latitude/longitude grid using the Barnes method. The DMW coverage is usually very good in the output layer but is of variable quality in the steering layer and lower layer. Thus, to improve the coverage, the global forecast system (GFS) winds are used as a low-weighted background field. The gridded horizontal winds are partitioned into nondivergent and irrotational components through calculation of the streamfunction and velocity potential. The diagnostics for TC intensity prediction include outflow layer divergence parameters and vertical wind shear of the nondivergent wind. Results show that these parameters estimated from the combined DMW and GFS fields provide information not contained in the GFS fields alone, especially those that are calculated from the irrotational (divergent) component of the wind field.

To test the impact on TC intensity forecasts, a six-year sample (2017-2022) of the DMW predictors for Atlantic and eastern and central North Pacific cases was used as input to a statistical intensity forecast model (DMW-TCIF) and a probabilistic rapid intensification index (DMW-RII). To capture the effects of other parameters not included in the DMW predictors such as SST and environmental moisture, the forecasts from NHC’s operational statistical hurricane intensity prediction scheme (SHIPS) and the companion SHIPS rapid intensification index (SHIPS-RII) were included as inputs to the DMW-TCIF and the DMW-RII. With this approach, the DMW predictions provide an adjustment to the operational SHIPS and RII models. The DMW intensity models are being tested on independent cases from the 2023 hurricane season and will be compared with the operational versions without the DMW input.

In addition, we will present a preliminary comparison of the operational DMWs with the experimental optical flow winds in the TC environment, to determine whether or not the statistical results can be improved with the new winds technique.