Dynamic Agricultural Land Unit Profile Database Generation using Landsat Time Series Images
American Geophysical Union Conference, San Francisco, CA, USA
Agriculture requires continuous supply of inputs to production, while providing final or intermediate outputs or products (food, forage, industrial uses, etc.). Government and other economic agents are interested in the continuity of this process and make decisions based on the available information about current conditions within the agriculture area. From a government point of view, it is important that the input-output chain in agriculture for a given area be enhanced in time, while any possible abrupt disruption be minimized or be constrained within the variation tolerance of the input-output chain. The stability of the exchange of inputs and outputs becomes of even more important in disaster-affected zones, where government programs will look for restoring the area to equal or enhanced social and economical conditions before the occurrence of the disaster. From an economical perspective, potential and existing input providers require up-to-date, precise information of the agriculture area to determine present and future inputs and stock amounts. From another side, agriculture output acquirers might want to apply their own criteria to sort out present and future providers (farmers or irrigators) based on the management done during the irrigation season. In the last 20 years geospatial information has become available for large areas in the globe, providing accurate, unbiased historical records of actual agriculture conditions at individual land units for small and large agricultural areas. This data, adequately processed and stored in any database format, can provide invaluable information for government and economic interests. Despite the availability of the geospatial imagery records, limited or no geospatial-based information about past and current farming conditions at the level of individual land units exists for many agricultural areas in the world. The absence of this information challenges the work of policy makers to evaluate previous or current government efforts for a given occurrence at the land unit level, and affecting the potential economic trade-off level in the area. In this study a framework is proposed to create and continuously update a land unit profile database using historical Landsat satellite imagery records. An experimental test is implemented for the agricultural lands in Central Utah. This location was selected because of their success in increasing the efficiency of water use and control along the entire irrigation system. A set of crop health metrics from the literature (NDVI, LAI, NDWI) is calculated and evaluated to measure crop response to farm management for its evaluation in time. The resulting land unit profile database is then tested to determine land unit profile groups based on land unit management characteristics. Comparison with essential inputs (water availability and climate conditions) and crop type (outputs) on a year basis is provided.
Torres-Rua, A. F. and M. McKee. 2012. Dynamic Agricultural Land Unit Profile Database Generation using Landsat Time Series Images, American Geophysical Union Conference, San Francisco, CA, USA, December 3-7, 2012, B43E-0447.