Space-time statistical analysis and modelling of nitrogen use efficiency indicators in China

Sandwich PhD supervised by Gerard Heuvelink (WUR, ISRIC), Ping He (CAAS, China) and Zhanguo Bai (ISRIC).

Nitrogen use efficiency (NUE) is significant for evaluating the utilization efficiency of nitrogen in agro-ecosystem and food production. It is especially crucial to balance crop yield, profitability and environmental sustainability for the strategies of sustainable development, and to establish crop nutrient application guidelines for different regions to achieve highly-efficient fertilizer management. In this project, we will analyse and model the spatial and temporal distribution of nitrogen use efficiency at provincial and county level according to related environmental explanatory variables. For this we will make use of stepwise multiple linear regression and compare the obtained models at different spatial scales. In addition, random forest regression will be used for exploring and modelling the NUE of wheat, maize and rice in China based on long-term experimental data. All model results will be statistically validated using k-fold cross-validation. Finally, an uncertainty propagation analysis of the county level model will be carried out and the contribution of uncertainty sources will be determined.