Monitoring the effects of Dracunculiasis remediation on agricultural productivity using satellite data

Dracunculiasis has been shown to be a major cause of agricultural work loss in many parts of West Africa. Monitoring the magnitude of agricultural loss caused by this parasite has been a problem due to the non-market nature of the agricultural output, which is often used for subsistence or barter. To test the hypothesis that 'Optimization of water and sanitation interventions in areas with high Dracunculiasis prevalence have a measurable socio-economic impact' a temporal analysis of Landsat Thematic Mapper (TM) satellite data was conducted. A region in which intervention had not taken place (the control) and a region in which it had were compared to determine the change in agricultural activity. Paired comparisons were made of the difference between two Landsat TM data sets: a 31 January 1986 scene (pre-intervention) and a 5 January 1991 scene (post-intervention) using the digital number (DN) values in bands 5, and 7, and the Normalized Difference Vegetation Index (NDVI) calculated from bands 3 and 4. The results of the paired comparison indicated that agricultural activity increased in regions where intervention had taken place. The analysis also indicated that the Harmattan winds may have effected the utility of the NDVI because of its differential effect on atmospheric scattering in the visible and infrared parts of the spectrum.