基于作物模型和多源遥感数据同化的农业干旱监测方法研究

The increasing global climate change has led to frequent drought disasters in China, posing a huge threat to agricultural production. Timely and accurate agricultural drought monitoring is the premise and basis for drought early warning and formulation of drought reduction strategies. How to effectively improve the estimation accuracy of drought monitoring indicators and ameliorate monitoring methods is a difficult issue to further strengthen the level of agricultural drought monitoring. With the rapid development of remote sensing technology and the deepening of data assimilation research, new ideas have been put forward to solve this problem. In this project, the typical wheat planting areas in China are research areas, and soil moisture is an indicator for agricultural drought monitoring. Soil moisture data assimilation schemes in rain-fed and irrigated agricultural areas are established by integrating crop model and multi-source remote sensing observations. On this basis, referring to response mechanism of crop water stress, an agricultural drought monitoring method combining crop root distribution and soil moisture data will be proposed. The drought characteristics and their regional differences in the past 20 years will be revealed, and the impacts of agricultural drought on crop yield will be analyzed. The research results can provide a scientific basis for improving the level of agricultural drought monitoring, reducing drought losses, and ensuring national and regional food security.

全球气候变化的日益加剧导致我国干旱灾害频繁发生，给农业生产带来巨大威胁。及时、准确的农业干旱监测是干旱预警评估、减灾策略制定的前提和基础。如何有效提高干旱监测指标估算精度、完善干旱监测方法是进一步提升农业干旱监测水平的难点问题。遥感技术的飞速发展和数据同化研究的不断深入使得解决这一难题有了新的思路。基于此，项目选取我国小麦典型种植区为研究区，以土壤湿度作为农业干旱监测的切入点，耦合作物模型和多源遥感观测构建雨养和灌溉农业区的土壤湿度数据同化方案；在此基础上，以作物水分胁迫响应机制为依据，提出融合作物根系分布和土壤湿度数据的新的农业干旱监测方法；揭示研究区近20年来的农业干旱特征及区域差异，分析农业干旱对作物产量的影响。研究成果可为提高我国农业干旱监测水平、降低干旱损失、保障国家和地区粮食安全提供科学依据。

在全球干旱灾害频发，农业生产受到广泛威胁背景下，提高干旱监测指标估算精度、完善干旱监测方法能够有效提升农业干旱监测水平，对于减轻灾害损失、保障粮食安全具有重要意义。本项目以我国粮食主产区黄淮海平原为研究区，结合土壤水量平衡方程和卫星遥感观测提出了农田灌溉参数的估算方法；在获取农田灌溉参数的基础上，利用参数优化法和顺序同化方法开展了灌溉农业区土壤湿度同化方法研究；考虑了作物根系分布及其对土壤水分吸收的影响，改进了现有基于土壤湿度的农业干旱指数；利用改进后的干旱指数分析了区域农业干旱特征及其变化趋势，探讨了农业干旱对作物产量的影响。研究结果表明：结合土壤水量平衡过程和微波遥感观测的估算方法能够较为准确地估算农田灌溉量和灌溉日期；耦合作物生长模型和遥感观测的数据同化方法能够显著提高土壤湿度估算精度，基于集合卡尔曼滤波的土壤湿度估算精度上要明显优于参数优化法的估算精度；考虑了根系分布对土壤水分吸收的影响，融合作物根系分布的土壤湿度指数能够改善农业干旱监测水平，提高干旱识别正确率；黄淮海平原发生的干旱主要以轻旱和中旱为主，干旱频率处于干旱中发区，北部地区的干旱频率要低于南部地区；总体上黄淮海平原农业干旱呈现加剧趋势，在冬小麦生长季内的次年春季表现尤其明显，对冬小麦产量影响较大的生育期主要出现在拔节、抽穗和灌浆期，应重点作物生长关键期的干旱风险防范。