基于多源数据的玉米干旱过程识别方法研究

The current studies on agricultural drought had paid more attention to a certain stage or the crop yields affected by drought, while neglected the research on the whole drought periods as it was a slow effect on crops. In this project, summer maize in Huanghuaihai Plain is selected, and the study is focused on the mechanism of drought evolution from the onset, development, to disaster/mitigation/recovery during tasseling - physiological maturity stage by integrating meteorological data, crop data, disaster data, remote sensing imageries, and RS-P-YEC model. Corresponding the relationship between soil relative humidity and crop drought, the main indicators of drought evolution, i.e. standardized relative air humidity index (SRHI), the normalized different water index (NDWI), and the standardized above ground biomass index (SBI), are used to identify the starting time suffering from drought stresses, the time and the intensity causing drought disasters, further to determine the thresholds at each stage during the drought evolution. The majority purpose of this project is to study the identification method of drought evolution based on the indicators with easy-to-obtain and good spatial continuity for summer maize during tasseling - physiological maturity stage, so as to provide a theoretical basis for monitoring drought evolution timely and accurately, avoiding the drought risk under climate change, and further putting forward to reasonable disaster prevention measures.

鉴于现行的农业干旱研究多关注干旱发生的某一阶段、或干旱对作物产量的最终影响，而疏于对干旱这种缓发性的灾害进行全过程监测，本项目以黄淮海夏玉米为主要研究对象，以夏玉米干旱易发时段(抽雄-成熟期)干旱过程为切入点，通过集成气象、作物生长发育、卫星遥感数据和农作物产量模拟模型，开展夏玉米干旱过程(发生、发展、成灾/缓解/恢复)不同阶段致灾机理研究。项目以标准化相对空气湿度指数、归一化差值水分指数、标准化地上生物量指数分别作为玉米干旱过程各阶段的主要指示指标，结合土壤相对湿度与农作物干旱的对应关系，判断玉米开始干旱的胁迫点、受旱的致灾点和临界致灾强度，厘定夏玉米干旱过程各阶段临界阈值，进而实现资料易于获取、空间连续性较好的黄淮海夏玉米抽雄-成熟期干旱过程识别方法研究，以期为及时准确地监测夏玉米干旱过程、规避气候变化新常态下可能的干旱风险、提出合理的农业气象灾害防御措施提供理论依据。

中国是世界上受干旱灾害影响最严重的国家之一，作为一个发展中的农业大国，干旱已经成为限制中国粮食生产的重要胁迫因子，严重威胁国家粮食安全。玉米是中国三大主粮作物之一，在夏玉米生育期内，气温高、蒸发量大、降水时空分布不均，易发生短期或阶段性农业干旱，造成玉米减产。本项目收集了研究所需的地面气象资料、农业气象资料、农业灾情资料，下载了相关卫星遥感产品，建立玉米干旱灾情基本数据集；基于降水量和相对湿度数据，计算标准化降水指数（SPI）和标准化相对湿度指数（SRHI），建立了基于旬、月尺度的SPI和SRHI的黄淮海平原夏玉米干旱灾害指示指标，并厘定不同时间尺度、玉米不同发育阶段干旱临界阈值，基于随机预留干旱灾害样本、典型站点长年代干旱样本对比分析旬和月尺度SPI和SRHI在识别玉米干旱方面的准确率，研究发现旬尺度的干旱指数在识别黄淮海平原夏玉米干旱时效果更好，SRHI较SPI在识别夏玉米干旱时更具优势；探索了基于卫星遥感地表反射率数据的玉米干旱灾害敏感性指标，发现基于近红外波段和对水分敏感的短波红外数据构建的归一化差值水分指数（NDWI）比基于红光-近红外波段数据构建的归一化差值植被指数（NDVI）在识别玉米干旱方面更敏感，且不同的短波红外波段在识别玉米不同发育阶段干旱时准确率差异较大；结合遥感-光合-农作物产量估算模型，评估了玉米长势的时间变化动态和空间差异，研究了不同降水年型下黄淮海平原夏玉米长势的差异，从东-西样带、南-北样带、西北-东南样带，揭示了夏玉米生长周期内降水量与长势的二次曲线关系；系统分析了黄淮海平原夏玉米不同发育阶段干旱时空分布特征，1961-2015年黄淮海平原夏玉米播种-抽雄期和抽雄-成熟期干旱发生频率分别为38%和25%，呈现出由南向北逐渐增加的空间分布特征，且64%年份中玉米播种-抽雄期的干旱发生范围大于抽雄-成熟期，表明黄淮海平原夏玉米在营养生长期更易遭受干旱胁迫。本研究可为及时准确地监测黄淮海平原夏玉米干旱灾害过程提供理论依据。