A Clear-sky Daily DLR product from the FY3D MERSI-2 observations

A daily clear-sky surface downward longwave radiation in the Arctic (60°N-90°N), produced from the FY-3D MERSI-2 thermal infrared (TIR) top-of-atmosphere (TOA) radiances, using the extremely randomized trees (ERT) machine learning algorithm. The dataset was firstly generated from the instantaneous MERSI-2 TIR radiances. The retrieval accuracy of the instantaneous LWDR dataset in RMSE and bias, validated with the Baseline Surface Radiation Network (BSRN) in situ measurements, are 14.14 W/m2 and 4.36 W/m2, respectively. Then, a diurnal model derived from the CERES SYN product was used to integrate the instantaneous LWDR data to daily product. The dataset has a spatial resolution of 1km.