Maruyama, Toshisuke ， Segawa, Manabu ， Takimoto, Hiroshi
2025-01-02 HYDROLOGICAL SCIENCES JOURNAL 2025 70(卷), 1(期), (162-175页)
Drylands cover approximately 40% of the Earth's land surface and account for up to 44% of cultivated lands. Therefore, evapotranspiration in drylands is an essential indicator of global warming because net radiation is partitioned into sensible (H, warming) and latent (LE, cooling) heat flux. This study aimed to estimate H (H-est) and LE (LEest) in arid regions using an inverse analysis (IA) method. The observed sensible (H-obs) and latent energy fluxes (LEobs) were compared with H-est and LEest over four time scales to confirm the reliability of the method, demonstrating that H-est and LEest were highly similar to H-obs and LEobs. Estimation accuracy was evaluated by the root mean squared error (RMSE) at each time scale, indicating that H-est and LEest can be reasonably estimated using IA. The method is expected to contribute significantly to estimating evapotranspiration for global warming research by providing for estimating H and LE in drylands.