Using time-series modis data for agricultural drought analysis in Texas

In this study time-series VCI data provided by George Mason University´s Global Agricultural Drought Monitoring and Forecasting System (GADMFS) are used for agricultural drought monitoring and forecasting. The validity of using VCI as a primary tool for drought monitoring is supported by the statistical result obtained in this article that the VCI is highly correlated with the PDSI at specific temporal and geospatial resolutions. Three classification schemes for drought severity are discussed here - Fixed Threshold, Natural Breaks (with Jenks) and Quantile schemes. Fixed Threshold Scheme is used though the article because it is computation effective compared with other two, and in the meantime the resulting map proves to be similar to the drought map provided by USDM. The correlation relationships between VCI and PDSI are not identical for different areas in Texas, depending on the vegetation distributions for the specific region. Areas with less variability in vegetation and fewer drought-resistant crops turn out to better reflect the correlations between VCI and PDSI.