A Physics-Constrained Implicit Profile Network for Continuous Reconstruction of Tropical Cyclone Near-Surface Wind Profiles [Dataset]
收藏资源简介:
This study uses quality-controlled tropical cyclone dropsonde observations collected by the NOAA Hurricane Research Division (HRD) during aircraft reconnaissance missions between 2001 and 2022. The dataset spans 93 tropical cyclones over the Atlantic and eastern North Pacific basins and includes storms covering a broad range of intensities, sizes, and environmental conditions. The original archive contains 13,058 records. Prior to model training and evaluation, all observations underwent the standard NOAA HRD quality-control procedures before analysis. Only records containing the required variables, including height, zonal and meridional wind components, storm center position, maximum storm wind speed, radius of maximum wind, and dropsonde relative location, were considered. The wind profiles were first converted to finite numerical arrays, and profile levels with missing height, zonal wind, or meridional wind values were removed. Duplicate height levels were merged by averaging the corresponding wind components. The analysis was then restricted to the near-surface layer from 10 to 300 m.



