Advances in ISCCP-based Surface Fluxes at Higher Spatial Resolution from the Surface Radiation Budget Project

The NASA Langley Research Center (LaRC) Surface Radiation Budget (SRB) project produced 3-hourly shortwave and longwave surface and top of atmosphere radiative fluxes for the 1983-2017 in its most recent version, Release 4 Integrated Product (IP) in collaboration with other GEWEX collaborators (Kummerow et al., 2019, Stackhouse et al., 2020, ATBD). This version uses the newly recalibrated and processed ISCCP HXS product as its primary input for cloud and radiance data, replacing ISCCP DX with a ninefold increase in pixel count (10 km instead of 30 km).  Previous work showed comparisons to BSRN and to ocean buoy measurements showed ensemble agreement for monthly averaged shortwave (SW or solar) wavelengths to be ~1 W m^-2 bias with an RMS of 14.7 W m^-2 RMS and longwave (LW or thermal infrared) ~+1 W m^-2 bias with a 15.9 W m^-2 RMS.  However, we also found that utilizing the Tselioudis (2020) weather state analysis with ISCCP to partition fluxes by cloud state over the BSRN sites showed that particular cloud states, such as the convective cloud state, showed much larger biases, particularly in the SW.

To address such issues, and to better resolve surface radiative flux spatial variability, this talk describes advances to the SRB inputs and algorithms towards the next release, referred to as LaRC SRB future Release 5. Since the resolution of ISCCP HXS is 10 km (excluding pixels within 25 km of coast lines), the ISCCP data products have enough sampling to grid cloud properties at the 0.5°x0.5°on a global basis.  In the shortwave, the Pinker-Laszlo lookup table approach with a forward call to the Fu-Liou radiative transfer model as modified by the CERES team (Rose et al., 2006).  In addition to being a proven radiation code, Fu-Liou allows the calculation of fluxes at different atmospheric levels and spectral bands, which will provide more insight into the surface radiation budget, its variability and attribution.  Updates to various inputs are described such as surface spectral albedos, emissivities, surface skin and near-surface temperatures, atmospheric profiles, and aerosols optical properties.  

This talk presents the results of early versions of the new products from grid boxes containing BSRN and ocean buoy measurement sites and compare these surface fluxes to the previous version and also to other prominent available data products in the literature.  Key regional differences over oceans and land are assessed to evaluate the changes in the resolving the flux variability.  Although the Tselioudis “weather states” are classified for a 1°x1° resolution, an initial flux partitioning is made at both full and a degraded 1°x1° resolution to assess the new algorithms under different cloud state conditions. The newer algorithms rely on the Fu-Liou based radiative transfer for both the SW and LW providing fluxes within the atmosphere and at the surface and for spectral band fluxes.