Submitted to: BARC Poster Day
Publication Type: Abstract Only
Publication Acceptance Date: 3/26/2012
Publication Date: 4/19/2012
Citation: Gao, F.N., Anderson, M.C., Kustas, W.P. 2012. Integrating multiple satellite data for crop monitoring. BARC Poster Day. 2012 CDROM. Interpretive Summary:
Technical Abstract: Remote sensing provides a valuable data source for detecting crop types, monitoring crop condition and predicting crop yields from space. Routine and continuous remote sensing data are critical for agricultural research and operational applications. Since crop field dimensions tend to be relatively small and crop condition evolves rapidly during the growing season, frequent satellite data at pixel resolutions of 30 to 60 m are required for many agricultural applications. Although there are a considerable number of remote sensing-based products available today, they have specific characteristics and are incompatible in terms of spectral bandwidth, spatial resolution, acquisition strategy, etc. For example, a satellite used in agricultural applications over the last ~30 years, Landsat, has a 30 m spatial resolution in the visible and near-infrared wavelengths and 16-day revisit cycle. The 30 m spatial resolution is good for crop monitoring at the sub-field scale, but a 16-day repeat acquisition is not frequent enough for monitoring conditions, especially when considering the reduction in useable data due to cloud cover. While Moderate Resolution Imaging Spectroradiometer (MODIS) provides daily coverage, it comes at price of having significantly coarser spatial resolution (250-500 m) which prohibits accurate mapping of crop type and condition of individual fields. We have been developing several techniques to integrate multiple sensors in a consistent way for crop condition monitoring. This poster will present three different techniques for the ongoing applications in: 1) spatial and temporal resolution integration; 2) leaf area index (LAI) product integration and 3) thermal data sharpening. Though integration of remote sensing data from multiple sensors, we are able to build consistent, frequent and high spatial resolution remote sensing data set. These approaches have been tested and validated in two agricultural experiment sites - the Soil Moisture Experiment of 2002 (SMEX02) in central Iowa and the Bushland Evapotranspiration and Agricultural Remote sensing EXperiment of 2008 (BEAREX08). Our approaches provide a feasible and cost effective solution for integrating remote sensing data from different satellite sources for crop type classification, crop condition monitoring and crop yield forecasting that will greatly benefit the USDA National Agricultural Statistics Service (NASS) and Foreign Agricultural Service (FAS) in many of their operational applications.