【作者】 华丽；

【导师】 蔡崇法；

【作者基本信息】 华中农业大学 ， 资源环境信息工程， 2013， 博士

【摘要】 土壤侵蚀是自然因素和人为因素共同作用的结果,土壤侵蚀分区是土壤侵蚀防治的基础。长期以来土壤侵蚀防治分区主要依据土壤侵蚀的自然因素,一定程度上忽视了社会经济发展等人类活动差异的重要性。在我国经济高速发展的今天,人类活动对土壤侵蚀的影响速度更快、频率更高、强度更大,人为因素的影响已不容忽略。如何从宏观层面客观准确地辨识土壤侵蚀特征及其驱动机制,分析自然与人为因素耦合作用下土壤侵蚀演化特点,以及依托自然环境及人类活动特点对土壤侵蚀影响的方向和程度,进行区划并制定防治措施具有十分重要的意义。本文基于宏观尺度,以社会经济因素为主并结合自然因素,构建区域尺度土壤侵蚀影响因素基础指标体系,基于偏最小二乘回归(PLS)方法构建模型选取影响土壤侵蚀的主导因子,定量分析区域土壤侵蚀及其影响因素的相互关系、空间分布特征和变化规律。以湖北省为例,剖析了人为因素时空变化对土壤侵蚀的影响特点,将分析结果应用于防治区划中,以期为区域水土保持治理区划提供科学依据。主要结论如下：1、湖北省土壤侵蚀空间分布特征表现为：全省不同时期均呈现山地侵蚀强度大、丘陵地区次之,平原区侵蚀强度较小的格局,与地形走势基本一致。1995年、2000年和2006年轻度及以上侵蚀面积依次减少,2006年轻度到强度侵蚀分布面积减少,但极强度和剧烈侵蚀面积共增加588km2；土壤侵蚀分布特征变化存在空间差异性,总体表现为山区土壤侵蚀状况有所好转,鄂西鄂东县(区)单元大部分侵蚀面积不同程度减少了5%-20%,强度下降,而丘陵平原区侵蚀趋于严重,部分县(区)侵蚀面积不同程度增加了5%-20%,平均模数增加1000t/(km2·a)以上。2、区域地形是影响土壤侵蚀的长期稳定因素,社会经济活动与土壤侵蚀受地形因素的制约,呈现与地形分布趋势较为一致的空间分布格局,人为因素在2006年对土壤侵蚀的影响大于2000年。利用整合后的多个社会经济指标作为人为因素,结合地形地貌、水系、降雨三个自然因素,构建土壤侵蚀多因素PLS模型,确定不同时间截面与土壤侵蚀显著相关的因素,进行空间自相关和热点分析,结果显示：显著相关的社会经济因素受地形因素的制约,地形因素、社会经济因素、侵蚀三者存在较为一致的空间分异规律。尽管地形因素仍是主要影响因素之一,但人为因素影响力逐年提升。2006年与2000年相比,地形因素重要性下降,人为影响因素的比重增大。3、短时间尺度内,受国家宏观政策调控的影响,人类活动是驱动土壤侵蚀变化的主导因素。表现为：(1)人为影响因素的指标增多,各因素变异程度具有空间分布差异性；(2)人为影响下,坡度差均方差较大的区域,短期内土壤侵蚀恶化速度小于地形平缓区域；(3)研究时段中,产业结构(第一产业比例)、经济发展水平(农村居民纯收入)、水土保持措施、水利措施及道路修建等指标是主要因素,对土壤侵蚀变化有显著影响。4、以地形和降雨因素为基础得到一级区划6个子区域。分别建立6个区域土壤侵蚀变化驱动因素模型,分区模型精度均高于全局模型,且主导因子与区域经济发展、产业结构、政策导向关系紧密。根据土壤侵蚀及社会经济因素的变化趋势与特点,划分出16个土壤侵蚀二级区,针对区域自然条件、社会经济与土壤侵蚀变化特征,提出修复与治理措施。总体表现为：抑制土壤侵蚀的因素是水土流失治理措施、植树造林、农村生活条件和能源结构的改善；加速土壤侵蚀的因素是产业(农业、工业)发展中不合理的资源开发利用、经济发展过程中道路修建和城镇建设等。分区表现为：秦巴山地区主要负面影响是第一产业发展;鄂渝山地区农产品种植、道路修建加剧土壤侵蚀；南阳大洪山地区的坡耕地加剧侵蚀,盆地区域种植业对土壤侵蚀呈现抑制作用；江汉平原地区主要为城镇化建设与工业发展引起的人为水土流失；桐柏山区,人为扰动大,种植、畜牧、养殖等产业发展加剧土壤侵蚀；大别山区人口增长和基础设施建设等因素导致土壤侵蚀强度增大；幕阜山九岭山地区治理成效好,但工农业比重大的地区如不注意水保措施的实施则会加剧侵蚀。本文研究具有以下创新点：1、首次从宏观层面,综合多项社会经济指标,分析了自然环境大背景下人类活动对土壤侵蚀的动态影响,并将分析结果应用于防治区划中。2、基于偏最小二乘回归模型(PLS)动态量化分析土壤侵蚀指标和自然环境、社会经济因素间的关系,不但克服了传统回归分析的缺陷,而且相较于专家打分选择影响因素更加客观、通用性更强。3、利用PLS的多重优势提取主要因素,结合地理加权回归方法(GWR)的空间局部分析优势将土壤侵蚀和影响因素在空间上的分异特征直观的表现出来是本研究的另一个特点。更多还原

【Abstract】 Soil erosion is the result of both natural forces and anthropogenic factors, soil erosion regionalization is the key to soil erosion prevention. The soil erosion regionalization and prevention have been based on natural forces for a long time, ignoring the impacts of the differences in human activities. Particularly, with the fast socioeconomic development in China, human activities, compared with natural forces, affect soil erosion faster, more frequently, and more intensively in a short period, and can no longer be ignored. Therefore, it is vital to characterize regional soil erosion and the driving forces, quantify the relationship between soil erosion and natural forces and human activities, evaluate the extent and the direction of the impacts of human activities on soil erosion, and provide measures to regionalize and prevent soil erosion based on the extent and the direction of the impacts of natural forces and human activities on soil erosion.This research evaluated the combining effect of environmental and socioeconomic factors at large scale, and established a set of basic indicators to soil erosion at regional scale. The driving factors for soil erosion were selected using the Partial Least Square (PLS) regression models. Furthermore, this research quantified the relationship between soil erosion and the impact factors, and analyzed the variations and the spatial patterns of soil erosion and the impact factors. This research, conducted in Hubei province, aims to analyze the impacts of spatial and temporal change of man-induced disturbance on soil erosion and apply results to regionalization and prevention of soil erosion, in order to provide scientific evidence for characterizing and controlling soil erosion at regional scale. Below are the conclusions from this research.1. The spatial distribution of soil erosion in Hubei province was in agreement with topography, with serious soil erosion in mountainous area, less serious soil erosion in hilly area, and least soil erosion in plains. Overall, in2006and2000, the area suffered from moderate and serious soil erosion significantly decreased in comparison with the soil erosion area in1995. The area of mild soil erosion decreased in2006, but the area of serious soil erosion increased by588km2. The spatial distribution of soil erosion was variable. Overall, the area of soil erosion decreased by5-20%in the western and eastern Hubei province, and the intensity decreased as well. However, the soil erosion in hilly areas was deteriorated, and the area of soil erosion increased by5-20%in some areas, and average modulus increased1000t/(km2-a) above.2. The regional topography was a long-term and stable factor for soil erosion, but the impacts of human activities on soil erosion was increasing. Socioeconomic activities and soil erosion were restricted by topography, and thus demonstrated an agreement with the spatial distribution of topography. The PLS model for the impact factors of soil erosion was based on the topographic, hydrologic, and precipitation factors and the human factor, which was the combination of multiple socioeconomic indicators. The factors related with soil erosion were evaluated for each time period. The spatial auto-correlation and hotspot analysis revealed that these socioeconomic factors were affected by topography, agreed well with the spatial distribution of topography and soil erosion. Although topography is a primary factor for soil erosion, the impact of human activities is increasing. Compared with2000, the importance of topography was decreasing in2006, but the impact of human activities was increasing.3. Human activity is the driving force for the change of soil erosion in a short time period due to the national macro-control policy.The number of human-activity indicators increased, and different factors demonstrated different spatial distribution. With the impact of human activities, in a short time period, soil erosion deteriorated faster in mountainous area with higher slopes than in hilly areas. During studying period, industrial structure (proportion of the primary industry), economic development (the gross income of rural residents), soil and water conservation, and road construction had strong impacts on the change of soil erosion.4. Hubei province was divided into six primary sub-regions according to topography and precipitation. Based on the properties of soil erosion and the driving forces in primary sub-regions, the sixteen secondary sub-regions were developed. Based on regional natural conditions and the change of socioeconomic characteristics on soil erosion, optimal situations and possible strategies for ecological restoration for the targeting area were proposed. The characteristics of soil erosion dynamics in sub-regions were analyzed, and the driving force model was built for each sub-region. I found that the accuracy of simulation of sub-regional model was better than the model for the whole Hubei province, and the driving factors were strongly correlated with regional economic development, industrial structure, and policies. Factors that can reduce soil erosion include soil and water conservation, afforestation, improvement of living standards and energy structure in rural area, while the inappropriate utilization of resources, construction, and urbanization can accelerate soil erosion. The deterioration of soil erosion in Qingba Mountainis was mainly caused by industrial development. In ErYu Mountainis area, the increase in cultivation of agricultural products and the construction of infrastructure accelerated soil erosion, In Nanyang Basin and Dahong Mountain area, growing agricultural products in Nanyang basin reduced soil erosion, but in Dahong Mountain area farmland cultivation accelerated soil erosion. What worth noticing is that the urbanization and industrial development caused the serious soil erosion in Jianghan Plain area. In Tongbo Mountain area, human disturbances, such as the increase in cultivation of animals and agricultural products as well as the increase in the income of rural residents, accelerated the soil erosion in this area.In Dabie Mountain, the increase in population and infrastructure construction accelerated soil erosion. In Mufo Jiulin Mountain area, the measures taken to control soil erosion were effective, but the soil erosion in areas with large industry and agriculture proportions might be deteriorated without effective soil and water conservation.The highlights of this research include:1. this study analyzed the impact of human activity dynamics on soil erosion at large scale and based on the combination of multiple socioeconomic indicators, and applied results to regionalization and prevention.2. PLS model quantified the relationships among the soil erosion indicators, environmental factors, and socioeconomic factors. This method not only overcomes the shortcomings of traditional regression analysis, but also is more objective and general than the expert scoring.3. Primary factors were selected using PLS model, and the differences in spatial distribution of soil erosion and the impact factors were illustrated using local analysis advantages of GWR.更多还原