Monitoring watershed drought conditions through the landsat imagery interpretation

Lake area is one of the most important hydrologie indicators of drought in the ApalachicolaChattahoochee-Flint (ACF) watershed. However, conducting a field survey of lake areas over a large region is usually infeasible and costly. The Landsat satellite imagery with moderate spatial, spectral and temporal resolutions has the capacity to quantify lake areas. This paper demonstrated two methods unsupervised classification and a-Modified Normalized Difference Water Index (MNDWI(a)) - to estimate the lake water area in ACF watershed. The estimated lake areas were used to correlate other hydrologie factors like precipitation. The results showed better performance of unsupervised classification than MNDWI. No correlation between accumulated precipitation and lake area fluctuation was detected, indicating more complex hydrologie mechanism existing in the large river basin.