Publications

Zhou, SG; Cheng, J (2023). A New Bottom-of-Atmosphere (BOA) Radiance-Based Hybrid Method for Estimating Clear-Sky Surface Longwave Upwelling Radiation From MODIS Data. IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 61, 5001825.

Abstract
Surface longwave upwelling radiation (SLUR) is a critical parameter for studying the water-energy balance between the land surface and the atmosphere. In this study, we proposed a new hybrid method for estimating the clear-sky SLUR from the bottom-of-atmosphere (BOA) radiance (new BOA-hybrid method) of the Moderate Resolution Imaging Spectroradiometer (MODIS). There are two key parts in the developed new BOA-hybrid method. First, a physical four-channel algorithm was developed to estimate the atmospheric terms (atmospheric upwelling radiance and transmittance), which were used to calculate MODIS BOA radiance. Second, a linear model linking SLUR and MODIS BOA channel radiances was constructed using a large number of samples generated by extensive radiative transfer simulations. In situ measurements from 27 sites in four flux networks were collected to validate the new BOA-hybrid method. The average bias and root mean square error (RMSE) were 1.65 and 16.23 W/m(2) for the new BOA-hybrid method, which was superior to the original BOA-hybrid method whose bias and RMSE were 3.52 and 18.51 W/m(2), and slightly better than the classical hybrid method that estimates SLUR using the top-of-atmosphere (TOA) radiance (TOA-hybrid method) whose bias and RMSE were -1.64 and 17.44 W/m2, respectively. The performance of the new and original BOA-hybrid methods was similar when the total precipitable water (TPW) vapor was less than 1.5 cm. However, under a large TPW (i.e., TPW > 3 cm), especially in the case of a large viewing zenith angle (i.e., VZA > 45(?)), the accuracy of the original BOA-hybrid method decreased significantly, while the physical four-channel algorithm could effectively improve the accuracy of atmospheric correction, and the RMSE of the SLUR estimated by BOA-hybrid could be reduced by more than 10 W/m(2). In summary, the developed new BOA-hybrid method can estimate SLUR accurately and has great potential to produce a long-term high spatial resolution environmental data record of SLUR.

DOI:
10.1109/TGRS.2023.3268674

ISSN:
1558-0644