研究生知识库

中国西南喀斯特地区石漠化生态灾难已经严重制约了当地经济和社会发展，亟需对喀斯特石漠化进行治理，而对其空间分布信息准确提取是必不可少的。针对现有喀斯特石漠化信息提取技术存在不能发生喀斯特石漠化范围提取不准确，不同空间尺度喀斯特石漠化指标不统一、综合研究缺乏，尚未形成统一的技术体系，且对于大区域喀斯特石漠化的时空演变分析较少、相关因子研究繁杂等问题，特别是考虑传统石漠化信息提取精度虽高，但提取时间成本、人力物力较大，而且会部分受到人为主观因素的影响，为实时、高效地服务于大尺度资源环境规划、石漠化综合治理，亟需基于传统石漠化信息提取方法，构建不同空间尺度喀斯特石漠化快速高效的石漠化信息提取方法。本文以典型喀斯特流域尺度-后寨河流域，县域尺度（凤冈、德江、思南三个县）以及贵州省为研究对象，基于高精度影像识别提取并剔除不能发生喀斯特石漠化范围，进而分别利用传统喀斯特石漠化信息提取方法、增加考虑地表反照率后改进的喀斯特石漠化信息提取方法以及基于植被覆盖度、岩石裸露率、坡度、土壤有机碳（SOC）等相关因子对石漠化的贡献率所建立的大尺度喀斯特石漠化信息快速提取方法，对流域尺度、县域尺度、省域尺度喀斯特石漠化信息进行提取及精度验证，揭示不同空间尺度上喀斯特石漠化的时空演变规律，同时探明喀斯特石漠化分布与其关键自然、人为驱动因子的相关关系，初步构建适用于不同空间尺度的喀斯特石漠化信息提取技术方法体系，为流域尺度、县域尺度、省域尺度喀斯特石漠化综合治理提供技术方法支撑，同时得出2005年至2015年间典型喀斯特流域时空演变数据和规律分析。研究结果表明：（1）针对不同空间尺度喀斯特石漠化信息提取方法体系，本文基于传统喀斯特石漠化信息提取方法和指标因子，通过分析相关因子地表反照率、坡度、SOC与石漠化间的关系，提出了适用于流域、县域、省域尺度的喀斯特石漠化信息提取方法体系，具体如下：① 基于流域尺度的石漠化空间分布信息遥感提取技术：可以考虑在流域利用高清影像计算岩石裸露率、植被覆盖度，并基于传统石漠化信息提取的方法进行流域喀斯特石漠化信息提取，精度较高，同时操作方法简单，便于实现。② 基于县域尺度的石漠化空间分布信息遥感提取技术：在进行县域尺度喀斯特石漠化信息提取时，受喀斯特石漠化提取数据量增加的影响，提取过程变得缓慢，提取效率降低。可以考虑使用影像进行监督分类、决策树分类等建立县域尺度土地利用分布图，并用其来剔除不能发生喀斯特石漠化范围以减少计算工作量。同时通过分析流域喀斯特石漠化信息提取结果与地表反照率之间的相关关系，引入地表反照率作为石漠化发生范围的检验和校正，经过精度验证分析，该方法在县域尺度具有较好的适用性，而且便于实现。③ 基于省域尺度的石漠化空间分布信息遥感提取技术：此外，对省域尺度喀斯特石漠化信息快速提取，本文基于县域尺度的石漠化信息提取结果建立了快速提取石漠化的模型，通过对贵州省喀斯特石漠化信息提取和结果验证，发现该方法在大尺度喀斯特石漠化宏观信息提取和监测方面具有处理简单、工作量小、快速高效同时基本满足精度需求等优点。（2）基于上述不同空间尺度喀斯特石漠化信息提取指标因子和方法，本文分别对流域尺度、县域尺度、省域尺度喀斯特石漠化信息进行了提取，其时空分布及演变存在以下特点：① 后寨河流域喀斯特石漠化呈现以下趋势：2005年至2010年后寨河流域喀斯特石漠化程度减轻、未发生变化以及加重的面积占比分别为8.3%、 50.7%、41%， 2005年至2010年后寨河流域喀斯特石漠化程度整体在恶化，在后寨河流域西南部分和中部极少区域喀斯特石漠化程度有所减轻，其余区域基本未发生变化。2010年至2015年后寨河流域喀斯特石漠化程度减轻、未发生变化以及加重的面积占比分别为57.6%、30%、 12.4%，在2010年至2015年后寨河流域喀斯特石漠化程度大部分区域都呈减轻状态，而且喀斯特石漠化程度减轻强度较大，在后寨河南侧大部分区域喀斯特石漠化程度都呈减轻状态，且强度较大。综上所述，在2005年至2015年间，后寨河流域喀斯特石漠化演变因贵州省2008年至2010年开始实施退耕还林还草等喀斯特石漠化综合治理工程而呈现先恶化，后减轻的状态，同时通过对该区域喀斯特石漠化分布与地表反照率、坡度的相关分析发现，地表反照率和坡度可以作为喀斯特石漠化研究的辅助表征因子和驱动因子。② 黔北三县喀斯特石漠化呈现以下趋势：2005年至2010年黔北三县喀斯特石漠化程度减轻、未发生变化、加重的面积占比分别为50.6%、18.9%、30.5%，黔北三县石漠化情况总体上是减轻状态，而且从2005年至2010年黔北三县的重度石漠化减少比较明显，除了在德江县边界部分等存在部分恶化，在思南县东南侧、凤冈县西南侧、德江县北侧等石漠化程度减轻都比较明显；2010年至2015年黔北三县石漠化程度减轻、未发生变化、加重的面积占比分别为33.6%、50.2%、16.4%，在2010年至2015年黔北三县喀斯特石漠化程度大部分区域都呈明显减轻状态，只在德江县中心部分存在石漠化程度加剧，而该区域为德江县人口聚集区，主要为人为活动影响。受贵州省喀斯特石漠化综合治理影响，在研究区域的尺度扩大后，喀斯特石漠化总体的演变趋势表征比较明显，而小区域喀斯特石漠化的时空演变趋势被弱化了，因而亟需对不同空间尺度喀斯特石漠化进行综合研究，以达到对喀斯特石漠化的有针对性的治理。而且通过本文研究发现，在和流域尺度采用同样空间分辨率遥感影像进行喀斯特石漠化信息提取的前提下，流域尺度上喀斯特石漠化的演变空间分布呈现小斑块儿状、区块儿状，县域尺度上喀斯特石漠化石漠化的演变则呈现比较细碎状态。③ 贵州省喀斯特石漠化呈现以下趋势：从2005年至2015年贵州省喀斯特石漠化整体呈现减轻状态，但是也存在一些区域喀斯特石漠化状态加重，主要是贵州省中部和南侧部分小区域。从2005年至2010贵州省喀斯特石漠化程度减轻、未发生变化、加重的面积占比分别为35.6%、38.5%、25.9%，2010年至2015年黔北三县石漠化程度减轻、未发生变化、加重的面积占比分别为36.5%、40.2%、23.3%，可以发现石漠化减轻和未发生变化的面积占比相对县域、流域尺度明显降低，而且通过对比2005年至2010年与2010年至2015年贵州省喀斯特石漠化时空演变发现，虽然喀斯特石漠化总面积在减少，但是存在部分区域石漠化好转后又恶化的情形，这与贵州省人均耕地占有量少，人地矛盾比较突出所造成的，因而对于石漠化的治理，人地矛盾的问题解决十分重要。（3）喀斯特石漠化的关键自然、人为驱动因子对其驱动特征：本文对喀斯特石漠化的关键自然驱动因子、人为活动驱动因子及相关表征因子进行了分析，通过本文研究发现喀斯特石漠化发生区域主要发生于2°— 22°的中等坡度之间、地表反照率为0.12 — 0.21之间，而且随着坡度增加潜在喀斯特石漠化、重度喀斯特石漠化发生率增高，轻度喀斯特石漠化和中度喀斯特石漠化发生率降低，随着地表反照率增加潜在喀斯特石漠化、轻度喀斯特石漠化发生率增高，中度喀斯特石漠化、重度喀斯特石漠化发生率降低。可以发现，喀斯特石漠化的分布与其地表反照率具有一定关系，可以考虑将地表反照率作为喀斯特石漠化提取的表征因子，同时，喀斯特石漠化的驱动因素坡度在喀斯特石漠化的分布上与人为活动协同控制、影响。除此之外，针对喀斯特石漠化的人类活动影响因子，本文选取人口密度和GDP空间数据对其与石漠化间的关系进行了分析，发现人口密度和GDP对于喀斯特石漠化也具有明显的相关关系，受人类活动密度与人类活动强度的影响，喀斯的石漠化的分布呈现明显的地带性。

The ecological disaster of karst rocky desertification in southwestern China has seriously restricted the local economic and social development. It is urgent to control karst rocky desertification in Karst areas, and accurate extraction of spatial distribution information is indispensable. Aiming at the problems existing in information extraction technology of karst rocky desertification, such as inaccurate extraction of the range of karst rocky desertification, inconsistent indicators of karst rocky desertification at different spatial scales, lack of comprehensive research, lack of a unified technical system, less analysis of spatial and temporal evolution of karst rocky desertification in large areas, and complicated research on related factors, etc., especially considering traditional karst rocky desertification. Although the accuracy of information extraction is high, the time cost, manpower and material resources of extraction are relatively large, and will be partly affected by subjective factors. In order to serve large-scale resource and environment planning and comprehensive control of karst rocky desertification in real time and efficiently, it is urgent to construct fast and efficient information extraction methods of Karst Rocky Desertification in different spatial scales based on traditional methods of information extraction of karst rocky desertification. This paper takes Houzhai River Basin, county scale (Fenggang, Dejiang, Sinan) and Guizhou Province as the research objects, extracts and eliminates the range of karst rocky desertification Based on high-precision image recognition, and then extracts the improved information of karst rocky desertification by using the traditional method of extracting information of karst rocky desertification and increasing the surface albedo. Based on the contribution of fractional vegetation coverage, rock bareness rate, slope and soil organic carbon (SOC) to karst rocky desertification, a fast extraction method of large-scale karst rocky desertification information was established. The information of karst rocky desertification at watershed scale, county scale and provincial scale was extracted and its accuracy was verified, and the time and space of karst rocky desertification at different spatial scales were revealed. Evolution law and the relationship between the distribution of karst rocky desertification and its key natural and man-made driving factors are also explored. A preliminary technological method system for extracting information of karst rocky desertification at different spatial scales is constructed to provide technical support for comprehensive control of karst rocky desertification at watershed, county and provincial scales. At the same time, typical karst rocky desertification from 2005 to 2015 is obtained. Spatial and temporal evolution data and regularity analysis of special watershed. The results show that:(1) According to different spatial scales of karst rocky desertification information extraction method system, based on traditional methods and index factors of karst rocky desertification information extraction, through analyzing the relationship between surface albedo, slope, SOC and karst rocky desertification, this paper proposes a method system of karst rocky desertification Information Extraction suitable for watershed, county and provincial scales, as follows:① Remote Sensing Extraction Technology of spatial distribution information of karst rocky desertification based on watershed scale: high-definition image can be used to calculate rock bareness rate and fractional vegetation coverage in the watershed, and traditional extraction method of karst rocky desertification information can be used to extract karst rocky desertification information in the watershed with high accuracy, simple operation method and easy realization.② Remote Sensing Extraction Technology of spatial distribution information of karst rocky desertification based on County scale: when extracting information of karst rocky desertification at County scale, the extraction process becomes slow and the extraction efficiency decreases due to the increase of the amount of data extracted from karst rocky desertification. The land use distribution map at county level can be established by using image for supervisory classification and decision tree classification, and it can be used to eliminate the scope of karst rocky desertification which can not occur to reduce the computational workload. At the same time, through the analysis of the correlation between the extraction results of karst rocky desertification information and the surface albedo, the surface albedo is introduced as the test and correction of the occurrence range of karst rocky desertification. Through the accuracy verification and analysis, the method has good applicability in the county scale and is easy to realize.③ Remote sensing technology for extracting spatial distribution information of karst rocky desertification on Provincial scale: In addition, the information of karst rocky desertification on provincial scale is extracted rapidly. Based on the results of information extraction on County scale, this paper establishes a model for rapid extraction of karst rocky desertification. Through information extraction and result verification of karst rocky desertification in Guizhou Province, it is found that the method is macroscopic in large-scale karst rocky desertification. Information extraction and monitoring has the advantages of simple processing, small workload, fast and efficient, and basically meet the accuracy requirements.(2) Based on the above-mentioned index factors and methods for extracting karst rocky desertification information at different spatial scales, this paper extracts karst rocky desertification information at watershed scale, county scale and provincial scale respectively. The spatial and temporal distribution and evolution of karst rocky desertification information have the following characteristics:① The karst rocky desertification in Houzhai River Basin shows the following trends:From 2005 to 2010, the proportion of areas that have been reduced, unchanged and aggravated are 8.3%, 50.7% and 41% respectively. After 2005 to 2010, the degree of karst rocky desertification in the Zhaihe River Basin has worsened as a whole. The degree of karst rocky desertification in the southwest and central parts of the Houzhai River Basin has been reduced, while the other areas have not changed. From 2010 to 2015, 57.6%, 30% and 12.4% of the total area of karst rocky desertification in the Houzhai River Basin were reduced, unchanged and aggravated, respectively. From 2010 to 2015, the degree of karst rocky desertification in most areas of the Zhaihe River Basin was reduced, and the degree of karst rocky desertification was reduced more intensively, and the degree of karst rocky desertification in most areas in the Henan side of Houzhai River Basin was It is lightened and has high strength.In summary, from 2005 to 2015, the evolution of karst rocky desertification in Houzhai River Basin deteriorated first and then decreased because of the implementation of comprehensive control projects for karst rocky desertification, such as converting farmland to forests and grasslands from 2008 to 2010 in Guizhou Province. At the same time, through the correlation analysis of the distribution of karst rocky desertification with surface albedo and slope, it was found that the surface albedo and slope could be improved. As an auxiliary characterization factor and driving factor of karst rocky desertification research.② The karst rocky desertification in the three counties of northern Guizhou presents the following trends:From 2005 to 2010, the proportion of areas with reduced, unchanged and aggravated karst rocky desertification in the three counties of northern Guizhou was 50.6%, 18.9% and 30.5%, respectively. The situation of karst rocky desertification in the three counties of northern Guizhou was generally alleviated, and the reduction of severe karst rocky desertification in the three counties of northern Guizhou was more obvious from 2005 to 2010. Except in the border part of Dejiang County, there was some deterioration in the southeast side of Sinan County and Fenggang County. The degree of karst rocky desertification in the southwest side of the county and the north side of Dejiang county has been reduced obviously; the proportion of areas that have been reduced, unchanged and aggravated in the three counties of northern Guizhou from 2010 to 2015 is 33.6%, 50.2% and 16.4%, respectively. From 2010 to 2015, the degree of karst rocky desertification in most of the three counties in northern Guizhou has been significantly reduced, but only in the central part of Dejiang county, the degree of karst rocky desertification has been aggravated. However, this area is a population gathering area in Dejiang County, mainly influenced by human activities.Affected by the comprehensive control of karst rocky desertification in Guizhou Province, the overall evolution trend of karst rocky desertification is more obvious after the scale of the study area is expanded, while the spatial and temporal evolution trend of karst rocky desertification in a small area is weakened. Therefore, it is urgent to conduct comprehensive research on different spatial scales of karst rocky desertification in order to achieve targeted control of karst rocky desertification. Furthermore, based on the same spatial resolution remote sensing image at watershed scale, the spatial distribution of karst rocky desertification at watershed scale is patchy and patchy, while the evolution of karst rocky desertification at county scale is fragmentary.③ The evolution of karst rocky desertification in Guizhou Province shows the following trends:From 2005 to 2015, the karst rocky desertification in Guizhou Province showed a reduced state as a whole, but there were also some regions where the karst rocky desertification was aggravated, mainly in the central and southern parts of Guizhou Province. From 2005 to 2010, 35.6%, 38.5% and 25.9% of the areas with reduced, unchanged and aggravated degree of karst rocky desertification in Guizhou Province, respectively. From 2010 to 2015, 36.5%, 40.2% and 23.3% of the areas with reduced, unchanged and aggravated degree of karst rocky desertification in the three counties of northern Guizhou Province were respectively. It can be found that the areas with reduced and unchanged degree of karst rocky desertification accounted for significantly relative county and watershed scale. By comparing the temporal and spatial evolution of karst rocky desertification in Guizhou Province from 2005 to 2010 and 2010 to 2015, it is found that although the total area of karst rocky desertification is decreasing, there is a situation that karst rocky desertification deteriorates after improving in some regions, which is caused by the less per capita arable land occupancy and the more prominent contradiction between human and land in Guizhou Province. Therefore, for the control of karst rocky desertification, the contradiction between human and land is contradictory. Problem solving is very important.(3) The key natural and anthropogenic driving factors of karst rocky desertification:In this paper, the key natural driving factors, human activities driving factors and related characterization factors of karst rocky desertification are analyzed. It is found that the occurrence area of karst rocky desertification mainly occurs between the middle gradient of 2°-22° and the surface albedo is between 0.12-0.21. The occurrence rate of potential karst rocky desertification and severe karst rocky desertification increases with the gradient. With the increase of surface albedo, the incidences of potential karst rocky desertification and mild karst rocky desertification increased, while the incidences of moderate karst rocky desertification and severe karst rocky desertification decreased. It can be found that the distribution of karst rocky desertification has a certain relationship with its surface albedo. It can be considered that the surface albedo can be used as a characterization factor for the extraction of karst rocky desertification. At the same time, the slope of the driving factor of karst rocky desertification is controlled and influenced by human activities in the distribution of karst rocky desertification. In addition, in view of the impact factors of human activities in karst rocky desertification, this paper chooses the spatial data of population density and GDP to analyze the relationship between population density and karst rocky desertification. It is found that population density and GDP also have obvious correlation with karst rocky desertification in karst. The distribution of karst rocky desertification in Karst shows obvious zonality under the influence of human activity density and human activity intensity.