The polar regions are believed to be the most sensitive to changes in the climate system. The Downwelling Longwave Flux (DLF) is the dominant component of the surface energy budget in the Arctic during the winter months. During this time, it is expected that changes in the climate system will produce detectable changes in the DLF. The lack of surface measurements in the polar areas necessitates the development of satellite based methods to estimate the DLF. These methods use retrieved cloud properties and meteorological soundings from satellite observations in radiative transfer calculations to compute the DLF at the surface. Here we present an intercomparison of these algorithms using estimates of cloud base height to calculate DLF during nighttime conditions from the NASA TOVS Polar Pathfinder (Francis et al. 1998) dataset in the Arctic. The results of each algorithm are also compared with data surface observations from the LEADEX and CEAREX to identify the most accurate algorithm(s).
The algorithms compared in this paper are from Gupta et al. (1992), Frouin et al. (1988), Francis (1996), Marshunova (1966), Zillman (1972), and Maykut et al. (1973). The Francis algorithm gives the best results to date by inferring cloud base and accounting for the ice and snow surfaces. This method gave a bias of -1.09 W/m2 with a RMSD of 19.14 W/m2. Other algorithms displayed fluxes that were underestimated with biases ranging from 20-30 W/m2. The reasons for these descrepancies are due to: the specification of cloud base height and layers; spatial and temporal uncertainties in the co-location of the measured and the calculated fluxes; and errors in the soundings from TOVS.
All algorithms assume one cloud layer and have prescribed cloud bases with the exception of the Francis algorithm that infers the cloud base height. Observations show at times as much as three simultaneous cloud layers. Thus, we explore the use of a random cloud overlap method to better approximate the cloud base height in the Frouin method. The random overlap method is currently being tested for operation processing at NASA Langley Research Center for the World Climate Radiation Programme (WCRP)/ Global Energy and Water Cycle Experiment (GEWEX) Surface Radiation Budget (SRB) Program. The NASA Langley WCRP/GEWEX SRB program uses the global cloud information from International Satellite Cloud Climatology Experiment (ISCCP). We compare both cloud properties and DLF from the TOVS based algorithms to those quantities produced from the SRB program.