Drought loss assessment combining remote sensing and a crop growth model for maize in Yunnan Province, China

Severe drought occurs every 2-3years in Southwest China, especially in Yunnan Province. The sown area of maize in 2010 was 1.35 million ha, representing one of the most important crops with the largest cultivation area among all crops in Yunnan Province. Agricultural drought is a complicated process under the combined impact of many natural and human factors, including precipitation, crop growth, soil properties, and irrigation. In this study, a drought loss assessment (DLA) approach integrating remote sensing and a denitrification-decomposition (DNDC) model were proposed to estimate maize loss in Yunnan Province in 2010. A fishnet with a grid size of 5km x5km was used to divide the entire province into 15,562 units. Basic data for every unit, including the maize area, fertilization, irrigation, precipitation, temperature, soil properties, and background nitrogen content, were collected for the construction of a database. Remote sensing data were used to monitor crop growth in different periods. Daily maize growth and drought stress in all units were simulated based on metrological and management information using the DNDC model. Scenarios involving full irrigation or no irrigation were used for the comparison of maize yields and assessment of drought loss. The results showed that the proposed framework considering both drought development and crop phenology was an effective approach for the analysis of drought impact on crop yield. In 2010, the maize losses caused by drought were 3.2 million tons.