Evaluation of land performance in Senegal using multi-temporal NDVI and rainfall series

Time series of rainfall data and normalized difference vegetation index (NDVI) were used to evaluate land cover performance in Senegal, Africa, for the period 1982-1997, including analysis of woodland/forest, agriculture, savanna, and steppe land cover types. A strong relationship exists between annual rainfall and season-integrated NDVI for all of Senegal (r = 0.74 to 0.90). For agriculture, savanna, and steppe areas, high positive correlations portray 'normal' land cover performance in relation to the rainfall/NDVI association. Regions of low correlation might indicate areas impacted by human influence. However, in the woodland/forest area, a negative or low correlation (with high NDVI) may reflect 'normal' land cover performance, due in part to the saturation effect of the rainfall/NDVI association. The analysis identified three areas of poor performance, where degradation has occurred over many years. Use of the 'Standard Error of the Estimate' provided essential information for detecting spatial anomalies associated with land degradation. (C) 2004 Elsevier Ltd. All rights reserved.