Accession Number : ADA332581

Title :   Retrieval Algorithms for Atmosphere Data Assimilation.

Descriptive Note : Final rept. 7 May-7 Nov 97,

Corporate Author : COMPUTATIONAL PHYSICS INC FAIRFAX VA

Personal Author(s) : Lumpe, Jerry D.

PDF Url : ADA332581

Report Date : 05 DEC 1997

Pagination or Media Count : 24

Abstract : Report developed under SBIR contract. This report details the development of algorithms for the purpose of assimilating multiple satellite remote sensing data sets of important geophysical parameters into instrument-independent three-dimensional gridded distributions. The assimilation problem has been formulated and solved as a general nonlinear retrieval problem, using the theory of optimal estimation. A detailed description of the method, and the specific structures resulting from its application to data assimilation, are provided. The algorithms have been tested on simulated satellite data sets for the specific problem of creating global ozone mixing ratio distributions from assimilation of satellite limb-viewing occultation and emission data. The results of these simulations clearly demonstrate the technical feasibility of the proposed approach. The potential applications of a general, rigorous data assimilation algorithm are widespread because of the increasing dependence on, and sophistication of, satellite remote sensing data in both the defense and civilian sectors. Examples include the suite of polar orbiting satellites operated by DMSP and NOAA which provide climatological data for operational weather prediction, multi-platform scientific missions such as NASA's planned EOS program, and commercial earth remote sensing programs such as LANDSAT and the French SPOT program.

Descriptors :   *ALGORITHMS, *ARTIFICIAL SATELLITES, DATA BASES, SIMULATION, EMISSION, OPTIMIZATION, PARAMETERS, STRUCTURES, WEATHER FORECASTING, FEASIBILITY STUDIES, REMOTE DETECTORS, INFORMATION RETRIEVAL, ASSIMILATION.

Subject Categories : Numerical Mathematics
      Unmanned Spacecraft

Distribution Statement : APPROVED FOR PUBLIC RELEASE