Geospatial tools for rainfed agriculture

Geospatial tools for rainfed agriculture


India’s Central Research Institute for Dryland Agriculture uses geospatial data and tools to identify and establish trends to promote rainfed agriculture

India is a vast country with total geographical area of 328 million hectares (Mha) and a net sown area of 142 Mha. Out of this, over 85 Mha is under rainfed cultivation which is the domain of Central Research Institute for Dryland Agriculture (CRIDA), a constituent of ICAR, Ministry of Agriculture, Government of India. The government has assigned top priority for developing rainfed agriculture under the XII Five Year Plan through the use of biotechnology and declaring a National Mission on Sustainable Agriculture.

CRIDA has research programmes to address various issues, including soil and water conservation through watershed development projects, water harvesting structures, study of agro-climatic situation, development of crops to withstand drought and shorter length of growing windows, soil fertility improvement, increasing carbon sequestration, increasing biomass availability for incorporation in soils, pest management and a host of others. Geospatial tools are used in a host of these research programmes.

Watershed-based development
The Watershed Development Programme is a major strategy for soil and water conservation in rainfed regions and is implemented across major agroclimatic or agro-ecological regions in the country. Changes in land use and land cover cause degradation, which requires restoration through interventions. Geospatial tools help in carrying out these tasks through land-use planning based on land capability and suitability, watershed-based development, soil and water conservation, locating water harvesting structures and farm ponds to improve water supply for rainfed agriculture, use of vegetation index to study crop condition, plant vigour, pest and diseases, soil fertility status, yield estimation etc, in addition to climate change studies and modelling.

Watershed-based development is critical for the fragile rainfed agroecosystems and hence a number of projects have been implemented since 1980s. During the XI Five Year Plan (2002-2007), it was felt that the guidelines for implementation of watershed projects required to be revised. This prompted a geomatics-based research study to develop methodology for assessing sustainability of watershed projects in rainfed agro-ecological-subregions in India. The CRIDA project undertook to develop a procedure for monitoring and evaluation of watershed projects using sustainability indicators to be measured by geospatial tools. Eight treated and untreated watersheds in Rangareddy and Nalgonda districts in the state of Andhra Pradesh were selected for monitoring and evaluation. While the usage of GIS, satellite data and GPS are routine features in delineation of watersheds and their development, their use for carrying out an objective post-facto monitoring and evaluation by a third-party after exit by a project implementing agency was new.

The study helped to identify 12 critical indicators for sustainable development of watershed projects. Spatial evaluation of the watershed projects indicated that in the selected districts, sustainable agriculture was being practised on 29-43% of the land in treated micro-watersheds. Rainfed agriculture in the untreated watersheds was found to be lagging, underlining the utility of watershed projects. Use of geomatics helped in developing an objective evaluation procedure in addition to measuring ‘sustainability’.

Effects of climate change
Sustainability of rainfed agriculture is threatened by climate change. ICAR carried out a study for assessing agricultural vulnerability at the agro-eco-sub-region and district level in the country using a vegetation index. An analysis of the Normalized Difference Vegetation Index (NDVI) time-series data showed variations, indicating the impact of climate change on vegetation growth and vigour. Vulnerable districts were identified for developing climate-resilient technologies. Variations in NDVI were correlated to standard precipitation index instead of the actual daily rainfall to study the impact of extreme weather events like drought, flood, heat and cold waves, cyclone, untimely rain, etc.

National Agricultural Statistics pertaining to agricultural production, yield and net sown area were analysed to corroborate results obtained from study of NDVI variations. All these information helped to identify agriculturally vulnerable regions in rainfed areas. It was found that over 92.98 Mha area in India experienced decreasing trend in NDVI while there was no change on 25.2 Mha. An increasing trend of NDVI was recorded on 183.96 Mha. Geographically decreasing trends in NDVI was noticed in the Western Ghats, Orissa and Chattisgarh regions of the country, the Northeast states and in lower Himalayas in Himachal Pradesh and southern Kashmir.

Overall there was an improvement in vegetation cover in the country. In 56 districts covering 30.93 Mha, a decrease in vegetative cover was registered, while in 41 districts with 22.25 Mha, there was no perceptible change in NDVI. In 457 districts accounting for over 249 Mha, a positive trend in NDVI was registered. An analysis of the Standard Precipitation Index indicated that while the regions of Deccan, West Bengal, Bihar, parts of northeast states, western Rajasthan and western J&K were receiving more rainfall, large parts such as the Indo-Gangetic Plain and Arunachal Pradesh were receiving less than normal precipitation.

Geospatial data and tools were fundamental in the above studies and helped identify and establish certain trends for the benefit of rainfed agriculture in India. At present, commercial software like ArcGIS and ERDAS Imagine are being used while an increasing need is being felt for the use of open-source GIS software and improved access to global and national datasets. While the National Remote Sensing Centre under ISRO is committed to supply free archived data, there is a need to develop tailored data like the GIMMS and MODIS datasets for public use in the country.