Assessing Impacts of Drought on Agriculture Production and Food Security in Karamoja of Northeastern Uganda with Meteorological and NDVI-Based Indices - Some findings and challenges

Drought monitoring and planning requires spatially and temporally continuous information and its impacts on production and on society can easily be quantified when long-term (both crop production and climate) information is available. Such historical information is scanty and at best qualitative for Karamoja in Northeastern Uganda, a region considered to be most vulnerable to drought. To demonstrate the capabilities of currently available satellite data in filling this data gap in smallholder agricultural regions; this study first characterized agricultural drought at multiple temporal and spatial scales using Normalized Difference Vegetation Index (NDVI) data (1999-2011) and monthly rainfall data (1960-2011). Correlation analyses of the NDVI based drought indicators, Standardized Precipitation Index (SPI) and a global product of Palmer Drought Severity Index (PDSI). Spatial information is derived for the 1999-2000 period using MODIS 250m resolution data. The 12 month SPI (SPI-12) had the highest correlation with the MODIS NDVI record from (1999-2011) derived indices reaffirming the cumulative effect of rainfall on vegetation during the growing season and the utility of NDVI as an indicator of drought. Time-series plots were generated, the droughts were ranked and spatial maps derived for the most severe droughts between 2000- 2011. Temporal drought information is correlated with proxy indicators such as food aid supplies, available historical production data, market prices from within and in neighbouring regions and with to primary data collected through interviews with farmers in Moroto district. This study demonstrates that operationalizing drought monitoring can be realized with remote sensing and further affirms the importance of drought and agriculture monitoring for food security.