戈壁表面砾石粒径遥感定量反演及其空间分异机制研究

Grain size of Gobi surface gravel could reflect the formation process of Gobi, we could learn about the Gobi characteristics, analyze the reasons of causing natural disaster by Gobi, cognize the laws of sand grain move and desert extension, and guide the execution of sand-prevention engineering by studying the grain size of Gobi surface gravel. The study would select the Ejina county in Inner Mongolia, China as experiment region, which belongs to a typical Gobi distribution region, using EO-1/HYPERION Remote Sensing imageries as the basic data resource, and adopting Remote Sensing, GIS, Spatial Information Statistics methods et al to study the spectral features of different grain levels of Gobi surface, quantitative estimation the grain size of Gobi surface gravel by remote sensing technique and spatial dissimilarity mechanism of Gobi gravel. Setting up the remote sensing estimation model of the grain size of Gobi surface gravel to solve quantitative estimation the grain size of Gobi surface gravel by remote sensing technique by analyzing the relationship between spectral of imageries, texture of imageries,roughness of imageries, geomorphology and the grain size of Gobi surface gravel. To clarify the effection degree and mechanism of the factors such as vegetation coverage, dry wet conditions of the Earth , desert varnish to the results of estimation the grain size of Gobi surface gravel. To clarify the spatial dissimilarity mechanism of grain size of Gobi surface gravel by adopting the model of Gobi surface grain size to analyze the spatial distribution pattern and collocation and by analyzing the geomorphology and external force. The study would provide the support for reforming and utilizing Gobi, reducing region ecological disaster, restoring and reconstructing the western economy.

戈壁表面砾石粒径尺寸反映戈壁形成过程信息，研究它可帮助了解戈壁特征，分析戈壁造成自然灾害的原因，认识沙粒迁移、沙漠扩展以及指导防沙工程。研究以戈壁典型分布区内蒙古自治区额济纳旗为实验区，以高光谱遥感影像数据EO-1/HYPERION为基本数据源，借助遥感、GIS、空间信息统计学等空间信息技术手段，开展戈壁表面砾石不同粒级遥感光谱特征，戈壁表面砾石粒径遥感定量反演以及砾石粒径空间分异机制研究。通过分析影像光谱、纹理、粗糙度以及地形地貌等特征与地表砾石粒径间关系，建立戈壁表面砾石粒径预估模型，解决应用遥感技术进行戈壁表面砾石粒径定量反演问题；弄清植被盖度、地表干湿度、荒漠漆等因素对砾石粒径反演影响的程度与作用机制；借助建立的粒径反演模型，从戈壁表面砾石不同粒径空间分布格局与空间配置着手，结合地形地貌以及外营力作用，弄清戈壁砾石粒径空间分异机制，为戈壁改造利用、区域减灾、西部经济建设服务。

研究在经过多次野外调查和专家论证基础上，首先把戈壁概念界定清楚，提出1套戈壁分类体系，基于分类体系制作了中国首幅戈壁分布图；在这些基础工作完成基础上，以嘎顺戈壁为实验区，通过采用戈壁地面高清影像，对高清影像进行模式识别，实现对戈壁表面砾石粒径进行自动精确测量，避免野外人工量测砾石粒径存在主观人为不确定性问题；通过野外采用ASD全波段光谱仪采集戈壁不同砾石粒径光谱基础上，分析光谱导数与吸收光谱特性，提取了对戈壁表面砾石不同粒级反应敏感波段，并将其应用到航天高光谱遥感影像上，实现了基于航天高光谱遥感影像的戈壁表面砾石粒径的精确估计；同时，研究还分别采用无人机高空间分辨率影像以及Landsat TM遥感影像数据，通过分析其影像光谱、纹理、粗糙度以及地形地貌等特征与戈壁表面砾石粒径间关系，建立了相应的戈壁表面砾石粒径预估模型，实现了对戈壁表面砾石粒径的定量反演；研究分析了植被盖度、砾石覆盖度等因素对戈壁砾石粒径遥感反演结果的影响；研究采用空间统计学方法，分析了戈壁表面砾石不同粒径空间分布规律，结合地形地貌以及外营力作用，揭示了戈壁砾石粒径空间分异机制。课题相关研究结果发表8篇学术论文，会议论文1篇；发布林业行业标准1项；获梁希科学技术奖三等奖1项；已培养2名博士生1名硕士生。