中文摘要：

研究分析了RCP2．6和RCP8．5气候情景下2021-2050年我国水稻高温敏感期（孕穗期一乳熟期）高温事件的变化趋势（较基准时段1961-1990年），并利用1981-2009年水稻田间观测资料，明确了水稻高温减产的主导因子，构建了主导因子与水稻减产率之间的经验回归关系式，在此基础上预估了未来RCP2．6和RCP8．5气候情景下我国水稻发生高温热害的风险变化。结果表明：①RCP2．6和RCP8．5气候情景下2021-2050年，全国各水稻种植区，水稻高温事件均有增加趋势，高温日数（HSD）、高温积温（HDD）都呈现增加趋势，高温持续日数（CHD）有延长趋势，其中华南双季稻区、长江流域单季稻区和东北单季稻区的HSD和HDD的变化幅度较为明显。②中国水稻高温中心在1961-2000年主要集中于湖南北部，与湖北省交界处附近，2种情景下2021-2050年均出现了向东北方向移动的趋势。③除东北区外，我国其余水稻种植区，孕穗一乳熟阶段的日最高气温连续3d超过35℃以上的有效积温HDD2是导致水稻减产的第一大要素，两者之间具有显著线性负相关关系；而东北区水稻产量更易受到孕穗一乳熟阶段的单日日最高气温超过32℃的有效积温SDD2的影响，且两者呈现出显著一元二次曲线关系。④与1961-1990年相比，2种气候情景下2021-2050年我国水稻发生高温热害的概率增加较大的地区，主要集中在长江流域单季稻区的湖北和安徽的大部分地区，华南双季稻区的广东、广西、海南省的大部分地区以及东北单季稻区的南部。

英文摘要：

This study investigated the changes of high temperature events during important growing period of rice （graining filling to maturity） of 2021-2050 due to climate change. Future climate scenarios were HadGEM2-ES simulation with RCP2.6 and RCP8.5 emission pathways. Relationship between high temperature and yield change was established from historical weather and field observations during 1981-2009 period. The impacts of high tem- peratures on China＇ s rice production up to 2050 were assessed by applying deduced regression models to climate scenarios. Key messages drawn from this exercise include： （1）High temperature event exhibited gradual increase from 2021 to 2050 under both RCP2.6 and RCP8.5 scenarios, characterized by increased number of high tempera- ture days （ HSD）, rising accumulated temperature with Tmax greater than 35 ℃ （ HDD）, and increased lasting days of high temperature （CHD）. The HSD and HDD increased substantially in double rice cropping system of South China, single rice cropping system of Yangtze River Basin and rice area of Northeast China. （2）High temperature hotspot was located near the border between Hunan and Hubei during 1961-2000, and might move towards northeast in the period of 2021-2050. （3）Except the Northeast, China＇ s rice production suffered most from in- creased HDD during grain filling to maturity, indicated by significant negative and linear relationship between yield and HDD, whereas rice in Northeast China was subject to the increase of SDD during grain filling to maturity, with a significant and quadratic relationship between the yield and SDD. （4）Compared to the high temperature risks during 1961-1990, climate change would increase the risks in majority of the rice area, especially in Hubei and An- hui-the central portion of Yangtze River Basin rice area, Guangdong, Guangxi and Hainan-south China double rice area, and south part of Northeast China single rice area.