HR: 1340h
AN: IN43B-1191    [Abstracts]
TI: Computation Methods for NASA Data-streams for Agricultural Efficiency Applications
AU: * Shrestha, B
EM: bijay@gri.msstate.edu
AF: GeoResources Institute, Mississippi State University, 2 Research Blvd., Starkville, MS 39759, United States
AU: O'Hara, C G
EM: cgohara@gri.msstate.edu
AF: GeoResources Institute, Mississippi State University, 2 Research Blvd., Starkville, MS 39759, United States
AU: Mali, P
EM: preeti@gri.msstate.edu
AF: GeoResources Institute, Mississippi State University, 2 Research Blvd., Starkville, MS 39759, United States
AB: Temporal Map Algebra (TMA) is a novel technique for analyzing time-series of satellite imageries using simple algebraic operators that treats time-series imageries as a three-dimensional dataset, where two dimensions encode planimetric position on earth surface and the third dimension encodes time. Spatio-temporal analytical processing methods such as TMA that utilize moderate spatial resolution satellite imagery having high temporal resolution to create multi-temporal composites are data intensive as well as computationally intensive. TMA analysis for multi-temporal composites provides dramatically enhanced usefulness that will yield previously unavailable capabilities to user communities, if deployment is coupled with significant High Performance Computing (HPC) capabilities; and interfaces are designed to deliver the full potential for these new technological developments. In this research, cross-platform data fusion and adaptive filtering using TMA was employed to create highly useful daily datasets and cloud-free high-temporal resolution vegetation index (VI) composites with enhanced information content for vegetation and bio-productivity monitoring, surveillance, and modeling. Fusion of Normalized Difference Vegetation Index (NDVI) data created from Aqua and Terra Moderate Resolution Imaging Spectroradiometer (MODIS) surface-reflectance data (MOD09) enables the creation of daily composites which are of immense value to a broad spectrum of global and national applications. Additionally these products are highly desired by many natural resources agencies like USDA/FAS/PECAD. Utilizing data streams collected by similar sensors on different platforms that transit the same areas at slightly different times of the day offers the opportunity to develop fused data products that have enhanced cloud-free and reduced noise characteristics. Establishing a Fusion Quality Confidence Code (FQCC) provides a metadata product that quantifies the method of fusion for a given pixel and enables a relative quality and confidence factor to be established for a given daily pixel value. When coupled with metadata that quantify the source sensor, day and time of acquisition, and the fusion method of each pixel to create the daily product; a wealth of information is available to assist in deriving new data and information products. These newly developed abilities to create highly useful daily data sets imply that temporal composites for a geographic area of interest may be created for user-defined temporal intervals that emphasize a user designated day of interest. At GeoResources Institute, Mississippi State University, solutions have been developed to create custom composites and cross-platform satellite data fusion using TMA which are useful for National Aeronautics and Space Administration (NASA) Rapid Prototyping Capability (RPC) and Integrated System Solutions (ISS) experiments for agricultural applications.
DE: 0430 Computational methods and data processing
DE: 0434 Data sets
DE: 0520 Data analysis: algorithms and implementation
DE: 0540 Image processing
SC: Earth and Space Science Informatics [IN]
MN: 2007 Fall Meeting