Evaluation of global gridded crop models in simulating sugarcane yield in China

Abstract

Sugarcane provides around 80% of the world's sugar production and is also one of the most efficient bioenergy crops. China is the third largest sugarcane-producing country in the world, and skillful simulation of the sugarcane yield in China is thus vital for global production and the trade of sugar and ethanol. Global Gridded Crop Models (GGCMs) are commonly used to predict global and regional crop yield as well as to assess the impacts of climate, environment, and agronomic management changes and feedbacks of crop growth. So far, two GGCMs (CLM5-crop and LPJmL) have been able to model sugarcane, but their performance in China remains unknown. In this study, the authors comprehensively evaluated the sugarcane yield simulations of these two models for the period 1980–2009 using the crop yield statistics collected from the National Bureau of Statistics of China. Results showed that these two models substantially underestimate the multi-year average sugarcane yield, with simulated yields less than one quarter of observations. In addition, CLM5-crop successfully simulates the spatial pattern, while LPJmL does not. In terms of temporal variability, the two models can reproduce the significant upward trend for the national average and in most provinces, but underestimate the magnitude. They also fail to simulate the pattern of interannual variability. The two models underestimate the sugarcane yield partly because they incorrectly set the sugarcane yield coming from the grain as other grain crops, which is inconsistent with the fact that the stem is harvested.摘要中国是全球第三大甘蔗生产国, 中国甘蔗产量模拟可服务于全球食糖和乙醇的生产和贸易. 全球格点作物模式CLM5-crop和LPJmL已实现对甘蔗的模拟, 但它们在中国的模拟能力未知. 本文评估结果表明: 两个模式均严重低估了甘蔗产量, 模拟均不足观测的1/4. CLM5-crop能有技巧地模拟产量的空间分布特征, 而LPJmL不能. 两个模式均不能合理模拟产量的年际变化, 且低估了产量的上升趋势. 模式低估甘蔗产量的部分原因是模式假设收割的是甘蔗的穗而非茎.