Monitoring hydrologic drought dynamic by using different indices based on remote sensing on SPOT-Vegetation images and ground data, case study: Marvdasht region, Fars Province, Iran.

In order to evaluate the capability of SPOT-Vegetation data for hydrologic drought monitoring in Marvdasht plain in Fars Province as a typical semi-arid region, a study plan was designed involving the production of Normalized Difference Vegetation Index (NDVI) correlating to Standard Water level Index (SWI) and precipitation data. SPOT Vegetation images were collected from 2001 to 2007 and processed. Geometric and radiometric corrections were performed and ten-day maximum NDVI maps were produced using ENVI software. Water level of 39 wells and precipitation data from 3 synoptic stations were also collected. The study covered a seven-year time period with three consecutive months in the growing season. Pearson correlation was performed to correlate NDVI values to SWI and precipitation data. In different time lag, the highest Pearson correlation coefficients (r values) were obtained between NDVI and SWI in the same month. No agreement was observed between NDVI and precipitation because unlimited using of aquifers for irrigation. Good correlations were obtained between average NDVI and average three-month SWI. The results indicated that SPOT-Vegetation derived NDVI well reflects SWI fluctuations in the study area promising a possibility for early hydrologic drought warning necessary for drought risk management.

Keywords:

Remote sensing , NDVI , SWI , Precipitation , Semi arid , Time lag , aquifer