Abstract
Land surface emissivity is important in calculating upward longwave
radiation from the surface. As a result of historical lack of observations land surface emissivity is currently set as constant in land surface
models (for example, 0.96 in CLM and 1 in NCEP) while, in the real world,
it varies with land type and season, with a value as low as 0.7 for some
desert regions. Our previous research illustrates that constant emissivity
in model induces an approximate error of 5-20Wm-2 into the surface energy
budget for arid areas. The primary goal of this project is to
(a) address
the uncertainty of the model's energy budget induced by constant land
surface emissivity in a coupled climate system, and
(b) reduce this uncertainty in NCAR CLM via satellite observations.Combining satellite observations into land surface model can
improve the model's initial and boundary conditions.
Earth
Observing System (EOS) MODerate Resolution Imaging Spectroradiometer (MODIS) observes the surface emissivity at 1km resolution using 6 spectral
bands. Before it can be used in CLM, MODIS spectral band emissivity must
to be converted to broadband and to be scaled up to model grid cell
resolution. In addition, we will compare MODIS emissivity with ASTER
observations for validation. A series of sensitivity simulations are
designed to examine the impacts of emissivity on the modeled sensib le
heat flux (SH), latent heat flux (LE), and skin temperature (Tskin).
Off-line NCAR CLM2.0 and coupled CAM2 with CLM2.0 (CAM2/CLM) will be used
in this work.This proposal complements our ongoing NASA EOSIDS project to
ensure we have adequate resources for this emissivity research. The
global emissivity data set developed and examined here will directly
improve NCAR community climate model and will be avai lable for other
modeling centers