Abstract The near‐surface specific humidity is critical for accurately estimating the enthalpy flux from the ocean, which plays an important role in tropical cyclone intensification. However, under the severe oceanic and atmospheric conditions of these storms, even spaceborne microwave radiometers struggle to retrieve reliable humidity data. This limitation stems primarily from heavy precipitation, which masks the water‐vapor signal in the observed brightness temperatures. To overcome these challenges, this study introduces a two‐pronged approach: (a) a neural network‐based screening of brightness temperatures to filter out precipitation‐contaminated measurements, and (b) the first use of low‐frequency microwave channel for humidity estimation in tropical cyclones. Comparisons with independent in situ observations demonstrate that our method provides more accurate near‐surface humidity estimates than previous techniques that do not use low‐frequency channels. These improvements will enhance the estimation of ocean‐to‐atmosphere enthalpy fluxes during tropical cyclones and may support better understanding of air‐sea interaction processes under extreme weather conditions.
Tomita et al. (Sun,) studied this question.