Connecting Agriculture Stress Index Systems at the Sub-National Level to the Next Generation of Seasonal Climate Forecasts: A General Approach to Transition from Monitoring to Forecasting.

Abstract:

Agriculture for food production remains a major contributor to the national economies of many developing countries. Often, these countries are characterized by agricultural landscapes that are heavily or even primarily dependent upon rainfall for crop irrigation and watering pastures for cattle. In the face of climate variability and change, decision-making processes at both the institutional and farm level are becoming more complex. Anticipating a potential agricultural drought and the associated impacts on food production would facilitate an informed risk-management strategy in climate-vulnerable agricultural landscapes. Systems for monitoring vegetation stress around the world have been successfully implemented at different geographical scales, and are used by leading global developmental and humanitarian agencies. Yet, these systems could benefit from the incorporation of a combination of seasonal (3-9 months) and sub-seasonal (2-6 weeks) forecasts, to transition from monitoring to a more proactive approach of forecasting agricultural droughts months in advance. This approach can, in turn, inform risk-management strategies at the farm and institutional level. The next generation of climate forecasts—hereinafter “NextGen”— developed by the International Research Institute for Climate and Society (IRI) and implemented by several National Meteorological Services around the world, opens new avenues for state-of-the-art research and applied science that has the potential to transform policy-making processes, and help local governments and developmental and humanitarian agencies achieve their goals. In this paper, we show the advantages of using a pattern-based-calibrated, multi-model ensemble, derived from the North American Multi-Model Ensemble (NMME), to forecast vegetation stress at subnational level and how this system can connect to all of the main agricultural monitoring systems worldwide based on the Normalized Difference Vegetation Index (NDVI). We then discuss the general approach that could be used to transform the current agricultural stress monitoring systems from one of monitoring of agricultural stress to one incorporating forecasts at temporal and spatial scales relevant to smallholder farmers, governments and humanitarian and developmental agencies.