Improved Algorithm to Estimate All-Sky Shortwave Net Radiation Based on Top-of-Atmosphere Albedo

Shortwave net radiation (SWNR) serves as the vital variable of radiative energy balance and plays a key parameter in global climate, hydrological, and land surface process models. SZA, PWC, DEM, AOD, TOA albedo and aerosol type are the key factors for estimating SWNR, and it is necessary to fully consider them. In the study, TOA albedo is estimated using MODIS data. An improved scheme is proposed for estimating SWNR by establishing a relationship between TOA albedo and SWNR based on SZA, PWC, DEM, and AOD parameters under different atmospheric conditions. The improved model is assessed using MOD TRAN simulation data, ground measurements, and comparative analysis with Wang-2024, Tang-2006, and CERES SSF product. The results demonstrate the superior theoretical precision of the improved scheme, based on MODTRAN simulation data, significantly outperforms existing methods, achieving bias and RMSE of less than 1 and 21 W/m² respectively. For rural aerosol, ground-based verification further revealed that the improved algorithm and Wang-2024 deliver superior accuracy for all-sky (bias<4.6 W/m<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup>,RMSE<82 W/m<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup>). Notably, the improved algorithm performed the highest accuracy for urban aerosol type (bias = 1.8 W/m², RMSE = 69.8 W/m²), effectively resolving the underestimation issue of Wang-2024 and overestimation by Tang-2006 and CERES SSF. Additionally, the improved algorithm demonstrates enhanced performance across varying AOD ranges. When AOD exceeds 0.7, the improved algorithm resolves the significant overestimation (25∼85 W/m²) of the existing algorithms and CERES SSF. For AOD values below 0.7, the improved algorithm maintains its superior accuracy. Furthermore, the improved algorithm enables more detailed and precise mapping of SWNR with higher spatial resolution. With advancements in theoretical accuracy and broader applicability, the improved algorithm is expected to serve a pivotal role in diverse application scenarios as remote sensing technologies continue to evolve.