区分复杂时空特征差异的干旱定量综合监测模型研究

项目名称： 区分复杂时空特征差异的干旱定量综合监测模型研究

项目编号： No.41501556

项目类型： 青年科学基金项目

立项/批准年度： 2016

项目学科： 地质学

项目作者： 周磊

作者单位： 北京建筑大学

项目金额： 20万元

中文摘要： 实时准确的干旱监测是预防和减轻干旱影响，减少旱灾损失的主要途径，是干旱研究的关键问题。干旱具有复杂的时空特征，目前干旱综合监测模型研究在干旱机理分析、模型因子选择、植被和地表温度信息的评价与集成、综合模型构建和验证方法方面仍然存在不足。本项目拟以黄淮海及周边地区为研究区，选择典型干旱事件分析干旱机理和植被与温度响应特征，利用干旱特征分类法构建干旱综合监测模型因子集；基于数据挖掘技术，精细区分旱灾的复杂时空特征差异，构建集成多模型和多指标且能全面定量刻画干旱特征的干旱综合监测模型；分别利用实测的旱灾灾情、土壤湿度和农作物产量数据验证模型的精度和效果，评价其在不同区域的适用性。实现区域干旱机制和发生发展特征的准确量化表达，为科学制定防旱、抗旱政策和措施提供依据。

中文关键词： 旱灾；遥感监测；机器学习；成灾机理；空间分析

英文摘要： As the key issue of drought research, accurate real-time drought monitoring is the main way to prevent drought, alleviate the influence of drought and reduce the losses. Drought has complex spatio-temporal characteristics. The current research of the integrated drought monitoring model is still insufficient in the field of drought mechanism analysis, the model factor selection, evaluation and integration of vegetation and land surface temperature information and integrated model construction and validation methods. Selected Huang-Huai-Hai and the surrounding region as the study area, this project will choose the typical drought event to analyze drought mechanism and the vegetation, temperature response characteristics to drought. Then the factor set of integrated drought monitoring model will be constructed using drought characteristics classification method. Based on data mining technology, the integrated drought monitoring model will be established by precisely differentiating complex spatio-temporal characteristics of drought. This model can integrate multi-model and multi-index, and can fully and quantitatively describe drought characteristics. Finally, the precision and effectiveness of the integrated drought monitoring model will be validated using field observed drought disaster, soil moisture and crop yield data. The regional suitability of the model will also be evaluated in different areas. This research will realize the accurate and quantitative expression for the regional drought mechanism and its development characteristics, in order to provide scientific basis for the policies and measures of the drought control and drought resistance.

英文关键词： Drought;Remote sensing monitoring;Machine learning;Mechanism of disaster;Spatial analysis