Center for Research on Environment and Water (CREW)
CREW's mission is to quantify and predict water cycle and environmental consequences of earth system variability and change through focused research investments in observation, modeling, and application.

Assimilation of MODIS and AMSR-E Land Products into the Noah LSM

P. Houser¹ (PI), X. Zhan², Y. Luo³

¹Center for Research on Environment and Water & George Mason University, Calverton, MD 20705

²University of Maryland Baltimore County, Baltimore, Goddard Earth Science and Technology Center, MD 21250

³Center for Research on Environment and Water, Calverton, MD 20705

Project Abstract

Satellite remote sensing data products can provide real-time information about the states and storages of the land surface that have been shown to enhance numerical weather and climate predictions.  For almost two decades, the efforts and investments of several federal government agencies (i.e.: NOAA, NASA, DoD) have fostered the space-based retrieval of numerous terrestrial data products.  Recently, global land data-products have been generated continuously, such as land cover classification (LC), vegetation index (NDVI), vegetation canopy leaf area index (LAI), fractional photosynthetically active radiation (fPAR), land surface temperature (LST), snow cover/depth (SC/SD), and land surface albedo (Alb) from NASA’s MODerate resolution Imaging Spectroradiometer (MODIS) on board Terra and Aqua satellites, and the snow water equivalent (SWE), surface soil moisture (SM) and surface temperature (ST) from the Advanced Microwave Scanning Radiometer (AMSR) on board NASA’s Aqua satellite and Japan’s ADEOS-II satellite.  However, almost none of these data products have been utilized in operational weather or climate prediction models.  One of the problems in the application of these data products is the lack of a well-tested approach to assimilate these satellite observations into a prediction system that is functioning in an operational framework.  To decisively accelerate the science of land surface satellite data assimilation in operational weather and climate models, we propose to develop and evaluate Kalman Filter data assimilation approaches to assimilate these MODIS and AMSR land data products into the Noah land surface model implemented in the North America Land Data Assimilation System (NLDAS) (Figure 1).

The work we propose directly addresses the program priority (4) Land Surface, listed in the NESDIS 2004-1 research announcement.  The proposed project tasks include the following: (1) Identify and incorporate relevant satellite land remote sensing data sets into the North America Land Data Assimilation System (NLDAS). (2) Derive and implement the observation functions (the forward radiation transfer models) needed in the assimilation of each of the satellite remote sensing products and implement the Kalman Filter data assimilation approaches with the Noah land surface model in NLDAS. (3) Examine the efficiency and benefit of assimilating the satellite data products into the Noah model and deliver the satellite data assimilation procedures to NOAA-NCEP.

The proposed work is based on the results of several current research projects: The North American Land Data Assimilation System (NLDAS) project which runs several land surface models (including Noah) using different forcing data sets, and the AMSR_E land data product validation project which has successfully implemented the Extended Kalman Filter to assimilate land surface soil moisture data into the MOSAIC land surface model.  This project will build on these research results by extending the Kalman Filter data assimilation technique to the assimilation of other land remote sensing data products with the operational Noah land surface model for operational weather and/or climate predictions.

                Figure 1

Project goals and objectives

 The primary goal of this proposed project is to enhance numerical weather and/or climate prediction models with the capability of assimilating satellite observations into their land surface model components in order to improve their prediction accuracy.  To reach this primary goal, we have to meet the following scientific objectives addressing the questions stated above:

1)   Identification and quality control of the continuously generated remotely sensed data sets of those land surface state or storage variables that are believed to have significant impact on numerical weather and/or climate predictions and are identical or directly related to the corresponding variables of the land surface model;

2)   Development of Ensemble Kalman Filter data assimilation approaches with the Noah land surface model within the North America Land Data Assimilation System (NLDAS) (see example results in Figure 2);

      3)   Examination of the efficiency and benefit of assimilating these land data products into the land surface model for numerical weather and/or climate predictions.

Figure 2

Last Updated: October 31, 2006