| 基于机器学习的参考作物蒸散量估算研究 |
| |
| 引用本文： | 毛亚萍,房世峰.基于机器学习的参考作物蒸散量估算研究[J].地球信息科学,2020,22(8):1692-1701. |
| |
| 作者姓名： | 毛亚萍 房世峰 |
| |
| 作者单位： | 1.中国科学院地理科学与资源研究所 资源与环境信息系统国家重点实验室,北京 1001012.中国科学院大学,北京 100049 |
| |
| 基金项目： | 中国科学院战略性先导科技专项(XDA20010302);国家自然科学基金项目(41971082);国家自然科学基金项目(U1503184) |
| |
| 摘 要： | 参考作物蒸散量（Reference Evapotranspiration, ET0）的准确估算对区域水资源管理和分配、流域水量平衡以及气候变化等研究具有重要作用。新疆地处我国西北干旱地区,水资源供需矛盾尖锐,精确估算该地区的ET0有助于其科学合理地调配水资源,缓解水资源供需压力。FAO推荐的Penman-Monteith法是计算ET0的标准方法,但该方法需要多项气象因子,而新疆地区气象站点较少且分布不均,精确完备的气象数据在新疆大部分区域难以获取。因此,如何使用有限气象因子获取高精度的ET0在新疆地区备受关注。本文基于中国气象数据网提供的新疆地区1980—2019年的地面气候资料日值数据集,在日和月尺度下,通过对最高气温Tmax、最低气温Tmin、平均气温Tavg、风速U2、相对湿度RH和日照时数n共6项气象因子进行敏感性分析,形成不同的气象因子组合;然后使用SVM、RF、GBDT和ELM 4种机器学习算法,以FAO-56 PM计算值为标准值,对新疆地区的ET0进行了拟合建模;最后,从拟合精度、稳定性和计算代价3个方面对模型进行评价。研究表明：① 在新疆地区,ET0对RH、Tmax和U2敏感系数级别为高,平均敏感系数分别为-0.516、0.283和0.266;n为中等,平均敏感系数为0.124;Tmin和Tavg为低,平均敏感系数分别为-0.016和-0.003;② 在日尺度,各算法在RH、Tmax、U2和n这4项气象因子为输入时精度较高（RMSE<0.5 mm/day,R2>0.95）,可对ET0进行精确估算;在月尺度,各算法使用RH、Tmax和U2这3项参数便可对ET0进行精确估算。SVM和GBDT这2种算法在日尺度和月尺度都有较好的适用性,可在相应尺度下使用较少气象因子替代FAO-56 PM公式对ET0进行估算。
|
| 关 键 词： | 参考作物蒸散量 机器学习 Penman-Monteith 新疆 支持向量机 随机森林 梯度提升树 极限学习机 |
| 收稿时间： | 2020-02-18 |
Research of Reference Evapotranspiration's Simulation based on Machine Learning |
| |
| MAO Yaping,FANG Shifeng.Research of Reference Evapotranspiration's Simulation based on Machine Learning[J].Geo-information Science,2020,22(8):1692-1701. |
| |
| Authors: | MAO Yaping FANG Shifeng |
| |
| Institution: | 1. State Key Laboratory of Resources and Environmental Information System, Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing 100101, China2. University of Chinese Academy of Sciences, Beijing 100049, China |
| |
| Abstract: | Accurate estimation of Reference Evapotranspiration (ET0) is essential to agricultural water management and allocation and hydrological cycle research. FAO-56 Penman-Monteith (FAO-56 PM) is the standard method to calculate ET0 recommended by Food and Agriculture Organization of the United Nations (FAO). But this method demands too many parameters and these meteorological inputs are not commonly available or unreliable, especially in Xinjiang province. Under this situation, machine learning algorithms have been introduced to estimate ET0 using fewer meteorological parameters and many comparisons of their prediction accuracy have been conducted. But the input combinations of meteorological factors are various and lack theoretical support. Meanwhile, the comparison of their performance at different time-scales has not been comprehensively conducted yet, and the good stability and less computational effort of models are also less to consider. The objective of this research was to evaluate machine learning algorithms' performance in modeling daily ET0 and monthly ET0 using fewer meteorological factors in Xinjiang. At this point, by using data collected from 41 weather stations in Xinjiang, this paper used Sensitivity Coefficient (SV) to evaluate the meteorological factors' influence degree to ET0 and then combined factors with high influence as input to Support Vector Machine (SVM), Gradient Boosted Decision Tree (GBDT), Random Forest (RF), and Extreme Learning Machine (ELM) in modeling daily and monthly ET0, and finally investigated and compared the performance of these algorithms from accuracy, stability and computational cost. The results showed RH (SV=-0.516), Tmax (SV=0.283)and U2 (SV=0.266) had high influence to ET0 followed by n (SV=0.124), while Tmin (SV= -0.016) and Tavg (SV= -0.003) exhibited low influence. In modeling daily ET0, models obtained satisfactory accuracy (RMSE<0.5 mm/day, R2>0.95) with input combination of RH, Tmax, U2 and n, while combination of RH、Tmax and U2 showed comparable accuracy for monthly ET0 prediction. The SVM and GBDT models showed the best performance and have been recommended for daily and monthly ET0 estimation in Xinjiang and maybe elsewhere with similar climates around the world. |
| |
| Keywords: | Reference Evapotranspiration machine learning Penman-Monteith Xinjiang Support Vector Machine Random Forest Gradient Boosted Decision Tree Extreme Learning Machine |
| 本文献已被 CNKI 等数据库收录！ |
| 点击此处可从《地球信息科学》浏览原始摘要信息 |
| 点击此处可从《地球信息科学》下载免费的PDF全文 |
|
