Tropical Cyclone Surface Winds From Aircraft With a Neural Network

• Published in: Journal of geophysical research. Machine learning and computation Vol. 2; no. 2
• Main Authors: DesRosiers, Alexander J., Bell, Michael M., DeHart, Jennifer C., Vigh, Jonathan L., Rozoff, Christopher M., Hendricks, Eric A.
• Format: Journal Article
• Language: English
• Published: Wiley 01.06.2025
• Subjects: machine learning; neural network; tropical cyclone
• ISSN: 2993-5210; 2993-5210
• DOI: 10.1029/2025JH000584

Estimates of the surface wind field in a tropical cyclone (TC) are required in real time by operational forecast centers to warn the public about potential impacts to life and property. In‐situ aircraft data must be adjusted from flight level to surface using wind reductions (WRs) since the aircraft cannot fly too low due to safety concerns. Current operational WRs do not capture all the variability in the TC surface wind field. In this study, an observational data set of Stepped Frequency Microwave Radiometer (SFMR) surface wind speeds that are collocated with flight‐level predictors is used to analyze the variability of WRs with respect to aircraft altitude and TC storm motion and intensity. The Surface Winds from Aircraft with a Neural Network (SWANN) model is trained on the observations with a custom loss function that prioritizes accurate prediction of relatively rare high‐wind observations and minimization of variance in the WRs. The model is capable of learning physical relationships that are consistent with theoretical understanding of the TC boundary layer. Radar‐derived wind fields at flight level and independent dropwindsonde in‐situ surface wind measurements are used to validate the SWANN model and show improvement over the current operational procedure. A test case shows that SWANN can produce a realistic asymmetric surface wind field from a radar‐derived flight‐level wind field which has a maximum wind speed similar to the operational intensity, suggesting promise for the method to lead to improved real‐time TC intensity estimation and prediction in the future. Plain Language Summary Extreme tropical cyclone (TC) winds can damage life and property. Aircraft provide useful data within TCs for forecasters to estimate the speed of damaging surface winds, but due to safety concerns much of the data comes from aircraft flight level several kilometers above the surface. Formulas for wind reductions (WRs) are used to reduce flight‐level wind to its expected value at the surface. However, the actual WRs can vary with TC wind structure and storm motion, such that current operational WRs are not capable of accurately accounting for these complicating factors. In this study, these factors are analyzed and the results are used to train a neural network (NN) model with observational data to predict a surface wind field from aircraft observations and other TC information. The model is called Surface Winds from Aircraft with a Neural Network (SWANN), and it prioritizes accurate predictions of high‐wind values given their importance to intensity estimation and damage potential. Validation of the model with data from past flights into TCs shows that SWANN improves on the current operational prediction formula. A test of a recent case shows SWANN is capable of producing accurate surface wind fields that can assist forecasters in real time. Key Points Tropical cyclone (TC) surface wind reductions from aircraft observations vary with flight altitude, TC motion, and intensity A neural network (NN) trained with aircraft observations is used to improve the reduction of TC winds from flight level to the surface The NN can predict observed surface winds with greater accuracy and better structural asymmetry than the current operational procedure