Integrated model for predicting rice yield with climate change
More details
Hide details
National Institute of Crop Science, Rural Development Administration, Mirayang-si, Gyeongnam, South Korea, 50424
Department of Rural and Agricultural Engineering, Chungbuk National University, Seowon-gu, Cheongju, Chungbuk, South Korea, 28644
Publication date: 2018-04-29
Int. Agrophys. 2018, 32(2): 203–215
Rice is the chief agricultural product and one of the primary food source. For this reason, it is of pivotal importance for worldwide economy and development. Therefore, in a decision-support-system both for the farmers and in the planning and management of the country’s economy, forecasting yield is vital. However, crop yield, which is a dependent of the soil-bio-atmospheric system, is difficult to represent in statistical language. This paper describes a novel approach for predicting rice yield using artificial neural network, spatial interpolation, remote sensing and GIS methods. Herein, the variation in the yield is attributed to climatic parameters and crop health, and the normalized difference vegetation index from MODIS is used as an indicator of plant health and growth. Due importance was given to scaling up the input parameters using spatial interpolation and GIS and minimising the sources of error in every step of the modelling. The low percentage error (2.91) and high correlation (0.76) signifies the robust performance of the proposed model. This simple but effective approach is then used to estimate the influence of climate change on South Korean rice production. As proposed in the RCP8.5 scenario, an upswing in temperature may increase the rice yield throughout South Korea.