EOS Validation Program

Institution: NASA Goddard Space Flight Center
                    Mail Code 614.3
                    Greenbelt, MD 20771
Phone: (301) 614-5772
FAX: (301) 614-5808
E-mail: Paul.Houser@gsfc.nasa.gov

Jared K. Entin, NASA Goddard Space Flight Center, Mail Code 614.3
James Foster, NASA Goddard Space Flight Center, Mail Code 614.3
Peggy E. O'Neill, NASA Goddard Space Flight Center, Mail Code 614.3
Jeffrey Walker, USRA/NASA Goddard Space Flight Center, Mail Code 614.3
Xiwu Zhan, UMBC GEST/NASA Goddard Space Flight Center, Mail Code 614.3

February 2003
April 2004

Provision of high quality remotely-sensed global land surface measurements is a key element of NASA's Earth Science Enterprise Program. However, it is recognized that these data fields will contain uncertainties due to imperfect instrument calibration and inversion algorithms, geophysical noise, representativeness error, communication breakdowns, and other sources. It is therefore essential that the accuracy and credibility of these remotely-sensed fields be evaluated for their use in critical research and applications.

Data assimilation systems have been used extensively in meteorology to expose significant defects in satellite data processing schemes, technology limits, bias, and noise. Modern data assimilation techniques use relevant prior data and a state-of-the-art computer model to estimate the state of the land surface. For each observation, a background value is derived from the model forecast for comparison. Systematic differences between observations and model predictions can identify systematic error, or identify uncharacteristically large differences in observations. Thus the consistency of the model provides guidance to identify observation problems in a data assimilation context.

We propose to determine the nature and variability of uncertainties in selected global soil moisture and snow products, measured by the EOS-AQUA AMSR-E sensor on a variety of time scales, and to analyze the effects of these uncertainties on the predictability of the global surface water and energy balance using land surface data assimilation techniques in near real-time. Specifically, we will use innovative land surface data assimilation techniques to check the quality control, physical consistency, and systematic realism of global EOS-AQUA land observations, in the context of the collaborative, real-time Land Data Assimilation System (LDAS) project. The goal will be to develop a nearly real-time operational land data assimilation system that will monitor the spatial-temporal AMSR soil moisture and snow observation quality, so as to provide feedback to mission operators of observational problems. This system will also extend AMSR-E products in time and space to produce consistent data assimilation land surface fields that will be valuable for use in subsequent analysis and application. Finally, selected in situ and airborne land surface observations will be used as a secondary test of data integrity. The proposed innovative validation strategies will feed back to improve the calibration and accuracy of EOS land-surface observations. These improvements will lead to enhanced characterization of the spatial and temporal dynamics of uncertainty in these critical land surface quantities and will benefit climate and weather prediction efforts.