Mapping high-resolution
land surface radiative fluxes from MODIS
Principal Investigator: Shunlin Liang1
Co-Investigators: Steven Running², John Townshend¹ & Si-Chee Tsay³
Collaborators:
Robert Wolf³, Crystal Schaaf & Alan Strahler
1University of
2 University
of
³NASA
Abstract
The surface energy budget (SEB) is the core of all land surface process models. It is the resultant of surface radiation and non-radiative components (e.g., sensible and latent heating). Land surface models usually rely on incoming radiation of the general circulation model (GCM) outputs as the atmospheric forcing, and calculate some key SEB variables (e.g., albedo and surface temperature) and the heating components. The NASA EOS program is routinely generating high-resolution albedo and temperature as the standard products that are used to validate the corresponding modeling components. There are currently several global radiative flux data sets derived from either satellite observations or GCM reanalysis, but the spatial resolution (usually >1°) and accuracy of these products are not satisfactory for high-resolution modeling and applications. The MODIS team has to disaggregate NASA’s 1°*1.5° reanalysis incident solar radiation to produce the 1km net primary productivity (NPP) product. There are also other compelling reasons that we urgently need to produce high-resolution radiative fluxes. For example, since it has been widely recognized that it is so difficult to separate emissivity and temperature from satellite observations, why do we not just produce upwelling longwave radiation instead?
Our overall objective is to develop the algorithms for routinely producing high-resolution land radiative fluxes from MODIS, including incident shortwave solar radiation (insolation) and photosynthetically active radiation (PAR), shortwave net radiation, longwave downward, upwelling and net radiation, and all-wave net radiation. All sky conditions for shortwave radiation and all-wave net radiation but only clear-sky conditions for longwave radiation products will be considered. Because of two MODIS sensors and the orbit convergence, both instantaneous and daily integrated products can be generated. The strategies for integrating multiple satellites (both geostationary and polar-orbiting) will also be explored.