中国农业气象

• 论文 • 上一篇    下一篇

多种格点作物模型对中国区域水稻产量模拟能力评估

孙擎,杨再强,杨世琼,王琳,赵和丽,韦婷婷,李佳帅,车向红,郑晓辉   

  1. 1. 南京信息工程大学气象灾害预报预警与评估协同创新中心,南京 210044;2. 南京信息工程大学江苏省农业气象重点实验室,南京 210044;3. 中国测绘科学研究院,北京 100830;4. 北京师范大学全球变化与地球系统科学研究院,北京 100875
  • 出版日期:2019-04-20 发布日期:2019-04-17
  • 作者简介:孙擎(1989-),博士生,主要从事作物模型、农业气象灾害及其风险研究。E-mail: sunqingmeteo@gmail.com
  • 基金资助:

    国家留学基金;江苏省研究生培养创新工程项目(KYLX16_0944)

Evaluations and Projections of Rice Yield from Multi-gridded Crop Model over China

SUN Qing, YANG Zai-qiang, YANG Shi-qiong, WANG Lin, ZHAO He-li, WEI Ting-ting, LI Jia-shuai, CHE Xiang-hong, ZHENG Xiao-hui   

  1. 1. Collaborative Innovation Center on Forecast and Evaluation of Meteorological Disasters, Nanjing University of Information Science&Technology, Nanjing 210044, China; 2. Jiangsu Key Laboratory of Agricultural Meteorology, Nanjing University of Information Science & Technology, Nanjing 210044; 3. Chinese Academy of Surveying & Mapping, Beijing 100830; 4. College of Global Change and Earth System Science, Beijing Normal University, Beijing 100875
  • Online:2019-04-20 Published:2019-04-17

摘要:

基于部门间影响模型比较计划(The Inter-Sectoral Impact Model Intercomparison Project,ISIMIP)FAST-TRACK 轮模拟中由5种国际耦合模式比较计划第五阶段(CMIP5)全球气候资料驱动下的 6 种水稻格点作物模型模拟水稻产量的结果,评估了格点作物模型模拟中国区域水稻历史产量(1980-2004年)的时空分布模拟效果,并基于多种作物模型等权重集合平均(Multi-Crop Models Ensemble,MCME)对未来(2020-2099 年)4 种不同典型浓度路径(Recommended Concentration Pathways,RCPs)情景下的中国区域水稻产量进行预估。结果表明:相对于单种水稻模型模拟的结果,采用MCME可以有效提高水稻模型在中国区域的模拟能力。MCME 模拟中国区域水稻历史年平均产量相关系数R和RMSE分别为0.798和1540.6kg·hm-2,在空间上对东北和西南地区模拟效果较好,其它地区模拟效果一般,模拟水稻产量的空间变率较大。未来随着气温和CO2浓度的上升,水稻产量呈增加趋势,在RCP8.5 情景下中国区域平均水稻产量在21世纪末增加最多,达到22%,RCP6.0情景下约增产15%,RCP2.6和RCP4.5情景下水稻产量在 21世纪上半叶增产,21世纪下半叶产量保持稳定甚至略有下滑,在21世纪末分别增产约4%和10%,在空间上东北和西南地区水稻增产较多,可达 40%以上,其它水稻主产区如长江中下游地区和华南地区增产较小。

关键词: 水稻, 格点作物模型, RCPs 情景, 多种作物模型集合

Abstract:

Based on The Inter-Sectoral Impact Model Intercomparison Project (ISIMIP) FAST-TRACK round’s results, we evaluated rice yield simulations of 6 global gridded crop models (GGCMs) driven by 5 Coupled Model Intercomparison Project Phase 5 (CMIP5) climate datasets from 1980 to 2004 as history period. Subsequently, the multi-crop model ensemble (MCME) was used to predict temporal-spatial distribution of future rice yield over China from the year 2020 to 2099 under different Recommended Concentration Pathways (RCPs). The results showed that MCME provided better performance for historic rice yield distribution with R of 0.798 and RMSE of 1540.6kg·ha-1 compared to single crop model results. MCME results showed better simulations in the north-east and south-westregions of China, but had poor performance in other regions. Moreover, the MCME overestimated spatial variability. Furthermore, under the increasing of temperature and CO2 concentration, rice yield had the largest growth of nearly 20% in the late 21st century under RCP8.5 scenario compare to the early 21st century and had a larger growth of 15% under RCP6.0 scenario approximately. For RCP2.6 and RCP4.5 scenarios, rice yield increased in the first half of 21st century, and stayed stable or even slightly decreased in the second half of 21st century, thus leading to a rise of only 4% and 10% respectively by the late 21st century. Rice yield would increase significantly (>40%) in the north-east and south-west regions of China. While other main rice planting areas like middle and lower reaches of Yangtze River and South China only experienced little increase.

Key words: Rice, Gridded crop model, RCPs scenarios, Multi-crop model ensemble