For cloudy-sky LW radiation, a strategy to combine the Land surface temperatures(LST) of MODIS and AMSR-E was proposed by considering the cloud coverage and the different surface penetration depth of the microwave and optical wavelengths. After fusing, a spatially continuous LST data was obtained and the corresponding surface LW upwelling radiation under all skies was then derived given the broadband emissivity. For longwave downward radiation (LWDR), a similar fusion method was employed to obtain atmospheric temperature and moisture profiles under cloudy-sky conditions by integrating MODIS and AIRS/AMSU atmospheric products. After that, the cloud thermal contribution and the LW contribution of sub-cloud layer were determined.
For cloudy-sky SW radiation, a look-up table was constructed based on numerus radiative transfer simulations, wherein, the cloud optical depth, cloud height and cloud phase were fully considered.
Finally, the all-sky SW and LW radiation was estimated by further combining the clear-sky radiation derived based on an artificial neuron network model.
The SURFRAD in situ radiation measurements, CERES and CloudSat products are selected to validate the derived surface downward and upwelling radiation under all-sky conditions. The verification results show that the newly developed strategies work rather well and derived all-sky radiation with better accuracy at a high resolution of 1-km.