Assessing the Impact of Spectral Surface Emissivity on Surface-Transmitted Irradiance Over the Sahara Desert

Surface-transmitted irradiance (STI), defined as the fraction of longwave radiation emitted by the surface that reaches the top of the atmosphere (TOA), remains an important area of uncertainty regarding the Earth’s radiation budget. Most climate models and radiative transfer applications either neglect surface emissivity effects by assuming blackbody emission or simplify the problem using broadband emissivity values. However, desert surfaces can exhibit spectral emissivities as low as 0.6 in the mid-infrared atmospheric window region, introducing potentially significant uncertainties in longwave radiative transfer calculations. This study investigates the local impact of spectral surface emissivity on STI over a Sahara Desert grid cell (20–30°N, 10–20°E) using line-by-line radiative transfer simulations performed with the Reference Forward Model (RFM). Surface emissivity inputs were derived from the CAMEL hyperspectral emissivity dataset, while ERA5 atmospheric profiles were used to describe atmospheric conditions. The experiment compared calculations performed using a graybody approximation based on broadband emissivity (BBE) against simulations using realistic spectral emissivity curves representative of different surface conditions within the selected grid cell. The results show that relatively small variations in BBE can correspond to substantially different spectral emissivity shapes, particularly within the 8–13 µm atmospheric window region. For the selected Sahara grid cell, STI values obtained using realistic spectral emissivity curves ranged from 85.7 to 96.82 W.m⁻², yielding a spread of approximately 11.1 W.m⁻² and a standard deviation of 3.1 W.m⁻² relative to the broadband emissivity estimate. The simulations indicate that the use of simplified emissivity assumptions can lead to local STI differences reaching up to 12% over the Sahara Desert. These findings demonstrate that broadband or blackbody emissivity approximations are insufficient for accurately representing longwave radiative transfer over arid regions. Incorporating realistic spectral emissivity information can substantially improve STI estimates and infrared satellite simulations, contributing to more accurate assessments of the Earth’s radiation budget and climate processes.