Smart and sustainable agriculture through weather and climate informatics and multi-scale modelling.

Abstract

Abstract Most agricultural processes, from choice of crops to sowing schedule to harvest management, depend on a number of weather variables; shifts in the weather regimes due to climate change determine crop variability at regional levels. Thus, advance and accurate information on weather variables, and their quantitative relations to the crop processes (agro-weather informatics) as well as quantification of role of climate conditions in agricultural processes (agro-climate informatics) can play critical to enabling roles in sustainable and smart (optimal in terms of planning, resource utilization and adaptability) agriculture. In particular, agro-weather-climate informatics have scopes and challenges including but beyond those in agriculture or weather informatics in isolation. This review examines principles and the scope of applications of weather and climate informatics in some of the key areas of agriculture such as crop choice, irrigation, energy management and long-term sustainability. While a serious bottleneck in application of forecast-based agro-advisories is low reliability of the forecasts, methodology combining optimization of forecast model, assimilation of data, multi-scale modelling and design of evaluation strategy for specific farm advisories is possible, so that the forecasts are risk-free and yet useful in spite of their errors. We provide a brief overview of weather and climate forecasting at different scales and the general methodology used; this is followed by an overview of status of these applications at different timescales such as daily, seasonal and beyond for a variety of issues such as precision agriculture and pro-active control of crop diseases. The roles of high-performance computing and multi-platform observations, including remote sensing, are outlined. This is followed by our conclusions and perspectives.