1 J. Farm Sci., 32(4): () 2019 REVIEW ARTICLE Applications of crop simulation models in global agriculture research: A review R. H. PATIL Department of Agricultural Meteorology, College of Agriculture, Dharwad University of Agricultural Sciences, Dharwad - 580 005, Karnataka, India E-mails: patilravi@uasd.in, ravipatil2005@gmail.com (Received: July, 2019 ; Accepted: October, 2019) Abstract: Models are built on the assumption that any given process and relation can be expressed in a formal mathematical statement(s) or equations, and they all can be inter-connected to run iteratively. Development of crop models started first in 1940s, but very few in the scientific community believed that the complex bio-physical and morphological processes of plant growth could be described mathematically, except possibly under controlled environments. In the early 1980s for the first time models were used as Decision Support System (DSS) tool. Soon after enhanced understanding of agro-ecosystem processes and synthesization of this to predict outcomes helped apply models to control and manage outcomes. The first product that came out of this effort was DSSAT model followed by APSIM model. Soon CGIAR centres started to develop and work with models. Development in modeling science and natural progression of scientific research went hand-in-hand. Since then models have been used in crop management, yield gap analysis, abiotic stress impact studies, potential yield, plant disease forecasting, inter-cropping and livestock-crop interaction studies. Soon models were used to predict climate change impact on food security and socio-economic issues at regional and global scale. Since 2010 modeling work expanded to multi-model inter-comparison, climate change and climatic variability impact assessment, quantifying uncertainties in models for further development, genomics studies, crop improvement programs, regional and global food security issues, GxExM and GxExMxS interactions, conservation agriculture (CA), agro-forestry and tree-crop-livestock interactions and tuber crops, and also as DSS tool for crop management under current and future climates. Key words: Crop models, Simulation, Prediction, CGIAR, DSSAT, APSIM Introduction Agriculture is a global enterprise practiced wide and far on all the continents, albeit in different climates, landscapes, agro- ecosystems, management systems and, of course, for different production systems, which to name a few, include grassland ecosystems for dairy and meat industry, arable cropping ecosystem for food grain and feed production, fruit and vegetable production systems, agro-forestry and tree farming system etc. However, when it comes to arable farming for food grain production, by far the largest number of countries on this Earth is engaged in food grain production system not only to feed their own citizens, but to export to feed global population. Post World War II the world we live in and the farming we practice has become more complex due to ever growing population and its demands for more food, water, and energy. Population on this planet has already crossed 7.3 billion and projected to cross 9.5 billion by 2050 AD, but land resources on this Earth are very finite and finding it difficult to meet the growing demand for food and fodder to people and animals. Increasing the food grain production and productivity per unit area to feed rapidly growing demands of the population, while protecting the natural resources and environment is easier said than done. This growing demand for food has to come from ever limited arable land; hence there is an increasing pressure on limited natural resources. This situation is further complicated by global warming and its impacts projected for coming decades (Wheeler and von Braun, 2013). But, despite all these challenges and odds against humanity, science needs to address these complexities. Therefore, one need to tread a careful path of balancing both while increasing global food basket, which requires a sustainable approach going forward. This requires new agricultural research to provide technologies and information to farmers, policy makers and other decision makers on how to achieve global food demand through sustainable agriculture systems across climates around the world. Why modeling? Research studies in the field of agriculture science are traditionally and predominantly carried out using conventional trial-and-error and experience-based field and/or laboratory based experimentations. In such studies crop performance and yield functions are derived from statistical analysis without referring to all the underlying bio-physical principles involved. However, use of correlation and regression analysis helps, to some extent, in qualitative understanding of the variables and their interactions, this indeed has enabled the progress of agricultural science (Jones et al., 2017). It is essential that ‘Science of Agriculture’ better explains and predicts growth of crops in managed and natural ecosystems in response to climate, soil and region and/or location specific management related factors. However, this requires quantitative prediction of complex interactions in a given system, which in turn depends on integration of information through different levels of organization in a system. This is possible through construction of statistical and simulation models. The first of the steps in simulation of crop growth begins with use and balance of carbon from its input for assimilation in the leaves which leads to the growth of vegetation. Such hundreds of steps are embedded in the system as webs and / or cycles of interactions. For instance, soon after placing a healthy seed into the soil its germination and emergence depends on