Sensitivity of the agroecosystem in the Ganges basin to inter-annual rainfall variability and associated changes in land use



The rate of growth in agricultural production has been decreasing in several regions of the world in recent years. The availability of water, which is one of the main inputs, is becoming limiting and more variable. In this article, we study the sensitivity of the agroecosystem to rainfall variability in order to identify vulnerable areas. We applied a longitudinal assessment of remote sensing time-series data, using the correlation between inter-annual rainfall anomalies and anomalies in Normalized Difference Vegetation Index (NDVI), a proxy for crop production. With a novel approach, we then identified whether the sensitivity results from a variation in crop growth or from a deliberate adjustment in the cropping pattern, reflecting a coping strategy. In our case study area, the Ganges basin, 25% of the basin area showed a significant correlation (p < 0.10) between rainfall and NDVI anomalies during the summer monsoon-dominated cropping season, both positive and negative. During the consecutive dry season, 18 % of the basin area showed a significant correlation, mostly positive. This variation in sensitivity shows the added value of spatially explicit information from remote sensing over lumped crop statistics. Primarily in the drier western part of the basin, a coping strategy of increasing fallow land in years with below-average rainfall was detected. Distinguishing a coping strategy from a crop yield reduction is important from both an economic and a hydrologic perspective.