Understanding the Propagation of GCM and Downscaling Uncertainty for Projecting Crop Yield: A Nationwide Analysis over India

Abstract:

General Circulation Models (GCM) play an important role in assessing the impacts of climate change at global scale; however, coarser resolution limits their direct application at regional scale. To understand the climate variability at regional scale, different downscaling techniques (such as dynamical and statistical) have been developed which use the GCM outputs as boundary condition to produce finer resolution climate projections. Although, both dynamical and statistical downscaling techniques have proven to be able to capture the climate variability at regional scale; there are certain uncertainties lying in their projections especially for a region like India which have complex terrain and climatic pattern. Here, the uncertainties, resulting from the use of multiple GCM and downscaling models, are quantified with the assessment of impacts on regional crop yield. Two crop models with different complexity-Decision Support System for Agro-technology Transfer (DSSAT) and Infocrop, are used, forced by dynamically (CORDEX, COordinated Regional climate Downscaling EXperiment) and statistically (Kannan and Ghosh, 2011; Salvi et al., 2013) downscaled data derived from multiple GCM’s. Advantage of these crop models is their ability to capture complexity of Indian condition. Yields of major crops in India, such as, rice, wheat and maize have been considered in the crop model and the impacts of climate change are assessed on their yields. The uncertainties in projected crop yields are also quantified, which must be incorporated for deriving vulnerability and risk maps for crop-climate assessments. This may further help to determine different crop management practices in order to reduce adverse impacts of climate change in future.