Temporal change in land use by irrigation source in Tamil Nadu and management implications.

Interannual variation in rainfall throughout Tamil Nadu has been causing frequent and noticeable land use changes despite the rapid development in groundwater irrigation. Identifying periodically water-stressed areas is the first and crucial step to minimizing negative effects on crop production. Such analysis must be conducted at the basin level as it is an independent water accounting unit. This paper investigates the temporal variation in irrigated area between 2000-2001 and 2010-2011 due to rainfall variation at the state and sub-basin level by mapping and classifying Moderate Resolution Imaging Spectroradiometer (MODIS) 8-day composite satellite imagery using spectral matching techniques. A land use/land cover map was drawn with an overall classification accuracy of 87.2%. Area estimates between the MODIS-derived net irrigated area and district-level statistics (2000-2001 to 2007-2008) were in 95% agreement. A significant decrease in irrigated area (30-40%) was observed during the water-stressed years of 2002-2003, 2003-2004, and 2009-2010. Major land use changes occurred three times during 2000 to 2010. This study demonstrates how remote sensing can identify areas that are prone to repeated land use changes and pin-point key target areas for the promotion of drought-tolerant varieties, alternative water management practices, and new cropping patterns to ensure sustainable agriculture for food security and livelihoods.

Remote sensing based change analysis of rice environments in Odisha, India.

The rainfed rice-growing environment is perhaps one of the most vulnerable to water stress such as drought and floods. It is important to determine the spatial extent of the stress-prone areas to effectively and efficiently promote proper technologies (e.g., stress-tolerant varieties) to tackle the problem of sustainable food production. This study was conducted in Odisha state located in eastern India. Odisha is predominantly a rainfed rice ecosystem (71% rainfed and 29% canal irrigated during kharif-monsoon season), where rice is the major crop and staple food of the people. However, rice productivity in Odisha is one of the lowest in India and a significant decline (9%) in rice cultivated area was observed in 2002 (a drought year). The present study analyzed the temporal rice cropping pattern in various ecosystems and identified the stress-prone areas due to submergence (flooding) and water shortage. The spatial distribution of rice areas was mapped using MODIS (MOD09Q1) 250-m 8-day time-series data (2000-2010) and spectral matching techniques. The mapped rice areas were strongly correlated (R(2) = 90%) with district-level statistics. Also the class accuracy based on field-plot data was 84.8%. The area under the rainfed rice ecosystem continues to dominate, recording the largest share among rice classes across all the years. The use of remote-sensing techniques is rapid, cost-effective, and reliable to monitor changes in rice cultivated area over long periods of time and estimate the reduction in area cultivated due to abiotic stress such as water stress and submergence. Agricultural research institutes and line departments in the government can use these techniques for better planning, regular monitoring of land-use changes, and dissemination of appropriate technologies.