Diurnal Variation of Cloud Physical Properties for Tropical Cyclones Over North Atlantic in 2019–2023

Abstract Diurnal cloud variations critically modulate tropical cyclone (TC) precipitation and intensity, yet remain poorly understood due to the lack of continuous observations. This study examined diurnal characteristics in North Atlantic TCs (2019–2023) based on retrieved cloud properties from GOES infrared measurements using ResUnet—a machine learning framework (ResUnet‐GOES). Results revealed distinct diurnal cycles: while average cloud top heights (CTHs) within the TC inner region (0–200 km) exhibited dual peaks at 05 and 14 local standard time (LST), cloud optical thickness (COT) and cloud water path (CWP) maximized singularly at 05 LST, contrasting with maxima in cloud effective radius (CER) at 16 LST. Notably, regarding deep convective clouds (DCCs) within the TC inner region, the average CTH, CWP, and coverage increased synchronously at night and peaked in early morning, providing the micro‐physical evidence for the early morning peak in TC precipitation. These diurnal signatures may serve as reliable indicators for diagnosing TC convective variability.