Du, Zhao Shiquan, Hu Zhichao, Wang
1 Introduction
Ecological risk is the possibility that an ecosystem is exposed to dangerous environmental conditions, in which regional ecological risk assessment focuses on the possible adverse effects of uncertain accidents or disasters in a certain area on the ecosystem and its components [1-3], including risk sources, risk receptors, exposure and response processes, ecological end points and other evaluation elements [with the increasing attention of decision makers, the risk receptors in ecological risk assessment have changed from key species, populations, communities and other natural ones. Risk sources tend to be complex and diversified accordingly [5-7], but from the mechanism of releasing pressure by risk factors, they can be roughly divided into different types, such as natural factors leading or human activities leading. At present, due to the deepening relationship between human activities and the natural environment, it is becoming an objective fact that human factors replace natural factors to lead environmental changes and regional development [8-9]. The types of ecological risks directly or mainly induced by human activities are often concerned, such as land destruction in mining cities [10], sediment enrichment in marine dumping areas [1 1], and wetland tourism development [65438]. In contrast, traditional natural disaster research pays more attention to the loss risk of people's lives and property [14- 16], and pays less attention to the ecological environment risk. At present, under the background of global environmental changes, geological disasters such as earthquakes, landslides and mudslides and extreme weather events such as low temperature, rain, snow and freezing occur frequently, which seriously damages the structure, function, safety and health of the ecosystem in the disaster area and greatly affects the natural foundation on which social economy, ecological civilization and human well-being depend [17]. Ecological risk of natural disasters, as an important branch of ecological risk assessment, defines the risk source as various disasters caused by natural events or forces, takes various natural ecosystems as risk recipients, and pays attention to the impact of disasters on human society. Carrying out ecological risk assessment of natural disasters not only enriches the ecological risk assessment system that emphasizes the study of man-made risk sources, but also expands new ways to manage and regulate ecological risks in combination with the characteristics of specific disaster-prone areas.
Regional ecological risk assessment is a process of studying the possibility and magnitude of negative ecological effects caused by some uncertain factors threatening the ecosystem in a large area [18]. As an integral part of ecological risk assessment, regional ecological risk assessment continues the principle and framework of the former, and the classic one is the "probability-loss" two-dimensional model [19], which focuses on the probability of ecological risk events and their possible consequences, respectively corresponding to the danger of risk sources and the potential losses of risk recipients when they suffer disasters; However, from the perspective of risk causal chain model [20], there is a lack of explicit expression of risk exposure response process from source to recipient. In fact, the "interface link" in which risk receptors are exposed to risk sources and trigger direct reaction is very important in the risk causal chain model, which implies the mutual feedback of regional ecosystem pattern and process, the change of ecosystem function, and the exchange state of material flow, energy flow and information flow between local disaster-stricken areas and their background environment [4]. The exposure response process reflects the sensitivity of risk receptors to losses when they are affected by risk sources, which can be characterized by the fragility of regional ecosystems. However, although the existing regional ecological risk studies express vulnerability and exposure to a certain extent, they pay little attention to it, and most of them regard vulnerability as a correction factor of loss [2 1-22], ignoring the importance of revealing the interaction between structure and function and the mutual feedback between pattern and process in vulnerability and exposure response.
The mountainous areas in southwest China have strong geological tectonic activities, complex topography and changeable climatic conditions, and are densely populated areas with high incidence of geological disasters such as landslides. At the same time, in order to alleviate the contradiction between man and land in the process of rapid urbanization, the construction and development activities of mountain towns in China are becoming more and more frequent [23-24], which is more likely to lead to the overload of mountain geological environment and aggravate the occurrence of geological disasters. Dali Bai Autonomous Prefecture in Yunnan Province (hereinafter referred to as Dali Prefecture) is a typical southwest mountainous area with frequent geological disasters. It is also a biodiversity hotspot area, an eco-environment sensitive area [25-26] and a pilot area for the construction and development of low hills and gentle slopes [27]. Therefore, the quantitative assessment of ecological risk of geological disasters in Dali Prefecture is particularly important for regional sustainable development. Based on this, this paper takes Dali as an example, paying attention to the risk source, risk receptor, exposure response process and ecological end point, and comprehensively measuring the ecological risk of landslide disaster on the basin scale based on the three-dimensional framework of "danger-vulnerability-potential loss", trying to integrate the relevant information of regional ecosystem structure and function, pattern and process into the risk assessment process by emphasizing the exposure response process, so as to enhance the understanding of regional ecological risk mechanism.
2 Overview of research areas and data sources
2. 1 Overview of research fields
Dali Prefecture is located at the junction of Yunnan-Guizhou Plateau and Hengduan Mountains in the west of central Yunnan Province, with a total land area of 29459 km2, with mountainous areas accounting for more than 90% (Figure 1). The topography of the whole state is high in the northwest and low in the southeast, with complex and diverse topography, especially along the Jinsha River, Lancang River, Nujiang River and Red River and their main tributaries. The large-scale distribution of "red layer" soft and hard interlayer makes the mountain slippery; The difference of neotectonic movement rose strongly, and the water erosion was strong; The three-dimensional climate is obvious, and the drought and flood seasons are distinct. The rainfall in the northwest is greater than that in the southeast, and increases with the elevation, forming seven rainy areas, such as Triple Mountain, Sepang Mountain, Diancang Mountain, Jizu Mountain, Wuliang Mountain, Laojun Mountain and Diaocaohoushan Mountain, with rainfall ranging from 2,400 mm to 2,500 mm, and rainfall in different places is concentrated and varies greatly, with an average annual rainfall of 1053 mm, with Yunlong being the most, Binchuan and Diaocaohoushan being the most. The area with heavy rainfall is the area with the most concentrated geological disasters in the whole state.
Graphic options
Figure 1 Distribution map of the study area and its geological hazards.
Dali Prefecture belongs to Jinsha River paraplatform and Sanjiang fold area, and the strata are exposed from Precambrian to Quaternary, especially Mesozoic. The main ore-controlling faults in Dali are: Chongshan West Side Fault, Lancang River Fault, Erhai-Honghe Deep Fault and Binchuan-Chenghai Fault. These four fault zones control the surrounding topography and lithology, and secondary faults are widely distributed in Quanzhou. There are frequent earthquakes in Dali prefecture, with strong earthquakes of magnitude 6 or above 16 times in 500 years. In recent years, moderate earthquakes of magnitude 4-6.5 have occurred frequently, causing serious secondary geological disasters. According to the summary of county and city planning data, the total number of registered geological disasters in Quanzhou in 20 10 was 779, including 5 12 landslides, 2 1 1 mudslides, 6 collapses, 37 unstable slopes, ground subsidence and ground fissures 13 (Figure
3 research methods
Ecological risk refers to the possible adverse effects of uncertain accidents or disasters on the ecosystem and its components in a certain area [4]. Based on the risk causal chain model [20], this paper identifies the risk receptor, exposure-response process and ecological end point, and constructs a three-dimensional framework of ecological risk assessment of "risk) = hazard) × vulnerability )× damage". Among them, "danger" focuses on the probability of landslide disaster, and is evaluated by information model based on geography, geology, human activities and other factors and their combinations. Vulnerability focuses on the vulnerability and sensitivity to disaster stress formed by ecological characteristics such as its own structure, which is the direct response of risk recipients to exposure to risk sources. This paper is characterized by ecosystem model. Loss is the "ecological end point" in the risk causal chain model, that is, the adverse ecological impact caused by disasters. The assessment is completed based on the quantitative accounting of various types of ecosystem services.
3.2 landslide risk assessment index system
According to the formation law and spatial distribution characteristics of landslides in Dali Prefecture, elevation (a), slope (b), lithology (c), normalized vegetation index (d), multi-year average precipitation (e), distance from river (f), earthquake density (g), distance from fault zone (h) and distance from human activities are selected respectively.
Graphic options
Figure 2 Spatial differentiation of landslide hazard indicators in Dali Prefecture Figure 2 Spatial differentiation of landslide hazard indicators in Dali Prefecture
Geographical environmental factors focus on the stability of disaster-prone environment, and select six indicators around topography, landform, soil, vegetation, meteorology, hydrology, etc.: ① From the disaster-causing mechanism, landslide belongs to gravity landform type, and elevation and slope directly affect the accumulation degree and distribution of loose accumulation layer and debris; (2) Precipitation is one of the important inducing factors of landslides, and the relationship between the number and scale of landslides and continuous rainfall and storm rain is very obvious. Precipitation infiltration will reduce the shear strength of soil and the friction resistance between soil and bedrock, increase the weight of soil and internal hydrodynamic pressure, thus inducing geological disasters; (3) Vegetation coverage, surface water transportation and erosion will all affect the stability of bank slope. Among them, NDVI is the most commonly used vegetation index in the study of estimating vegetation coverage by remote sensing, and it has a good linear correlation with the spatial distribution density of vegetation, so it is selected as one of the indicators in this paper. ④ The influence of surface water is characterized by the distance from major rivers. Rock structure determines the strength, stress distribution, deformation and failure characteristics of slope rock mass, which provides a material basis for the occurrence and development of landslide. In this paper, referring to the related research in southwest China [30], based on the geological map of China1:2.5 million, the information of rock types in the study area is extracted by vector data field fusion, and it is divided into five rock groups [30]( 1. Dolomite, thick rhyolite, etc. ; 2. Timely sandstone and siliceous conglomerate. 3. pyroclastic rocks, metamorphic basalts, etc. ; 4. Mud slate, coal seam, etc. ; 5. Clay, loose sediments, etc. ) (Figure 2c, 0 represents water body), similar rock groups have similar characteristics of landslide geological disasters. Geographical environment factors describe the key internal and external causes of landslide disasters, which are the direct basis of regional geological stability and have the greatest influence on landslide risk level, so it is given a higher weight in the three dimensions, which is set at 0.7703.
Geological structural factors mainly include two indicators, namely, the core density at the earthquake point and the distance from the fault zone. Among them, the influence of earthquake on landslide is mainly reflected in two aspects: first, landslide disaster occurs because the seismic force generated by earthquake directly acts on the rock and soil on the slope, and the damage accumulated on the fragile hillside may be affected by the previous earthquake in this area [31]; Secondly, earthquakes in mainland China are mainly controlled by active structures, and the environment prone to earthquake disasters is often extremely complex geological structures, developed faults and broken rocks, which indirectly affect the occurrence of landslide disasters. In this paper, the nuclear density of earthquake points is used to refer to the spatial agglomeration of seismic zone distribution. On the other hand, the influence of fault zone on landslide is that the distribution area of active fault is usually an area with strong differential movement, forming a variety of landforms such as valley depth, steep slope and large slope drop, which is prone to the sliding of bedrock or loose deposits, thus leading to landslide disasters. The occurrence of landslide disaster is controlled by the distance from the active fault zone. Using the buffer analysis function provided by ArcGIS, the buffer distribution map of active fault zone in the study area is generated. Geological structural factors directly or indirectly affect the occurrence of landslides through the current situation caused by historical geological activities such as earthquakes and fault zones, and the weight is set at 0. 16 18 among the three dimensions.
Human activities are also direct or indirect factors that aggravate the formation of geological disasters. Human engineering activities and economic development, such as village and town construction, agricultural activities, road engineering, mineral engineering and water conservancy projects, may affect rock and soil mass and aggravate slope instability. Vegetation destruction and ecological environment deterioration caused by land reclamation will also aggravate the activity intensity and scope of geological disasters such as landslides. In this paper, referring to the availability of data, the distance from the town and the distance from the highway are specifically selected to reflect the possible impact of human engineering activities and economic construction on landslide disasters. In Dali Prefecture, where the ecology is fragile, the rocky desertification is serious, and the construction of mountain towns is needed at the same time, electricity may induce human disturbance activities on a certain scale, but the impact on regional landslides is relatively small compared with geological structures, so the weight of human activity factors is the lowest, which is set at 0.0679.
4 Results and analysis
4. 1 landslide hazard factor information
4.3 Ecological Vulnerability and Potential Ecological Loss
Taking 367 three-level river basins as basic evaluation units, the comprehensive vulnerability index is calculated according to the natural breakpoint method and divided into five grades. The higher the level, the more serious the vulnerability (Figure 4a). Then, with the help of Moran's I index (significance level is 0.05), the spatial aggregation degree of ecological vulnerability of risk recipients is analyzed (Figure 4b). The results show that the high vulnerability areas of disaster-bearing bodies are distributed in four first-class basins in Quanzhou, mainly concentrated in the south of the Red River basin and the southeast of the Jinsha River basin where human activities are strong. The lowest level of vulnerability (1) is mainly distributed in the northern and eastern parts of Dali, which are less disturbed by human activities, and the low-value areas of vulnerability are often in a low-low aggregation state, such as Mijiang River Basin, Daoliu River Basin and the eastern part of Qingshui River Basin. At the same time, high-low-aggregation and low-high-aggregation basins are rarely distributed in Quanzhou, accounting for only 5.58% of the 6 1 third-level basins with significant spatial autocorrelation of ecological vulnerability. To a certain extent, this also confirms the holistic characteristics of the interaction of river basin unit ecosystems. Due to the relatively continuous attribute distribution and relatively low spatial heterogeneity of natural geographical elements such as geology, geomorphology and soil hydrology under large-scale background, and the flow of ecological processes in an open space, adjacent basins often show the similarity and vulnerability of ecosystem structure.
Graphic options
Fig. 4 Ecological Vulnerability Grade and Spatial Agglomeration Characteristics of Sub-basins in Dali Prefecture Fig. 4 Valley-level Economic Vulnerability and Spatial Agglomeration in Dali Prefecture
The grid layer of ecosystem services is standardized by extreme value range, and then the potential ecological loss of grid scale and watershed scale is obtained by spatial superposition and statistical calculation, and it is divided into five grades by natural breakpoint method. The higher the grade, the more serious the loss (Figure 5). The results show that the potential ecological losses in Quanzhou are mainly in junior high schools (Figure 5a). The fifth grade is mainly concentrated in Binchuan County and Heqing County, and the 1 grade is mainly concentrated in Erhai Lake, Jianhu Lake, West Lake and Yibihu Lake, mainly because forests mainly provide net primary production, water conservation and soil and water conservation services, and food supply is mainly provided by cultivated land, without focusing on water ecosystem services; In addition, the lower-level potential losses are relatively distributed in the northeast of Xiangyun County and the southwest of Yunlong County.
Graphic options
Figure 5 Potential Ecological Loss in Dali Prefecture Figure 5 Potential Ecological Damage in Dali Prefecture
As far as the basin scale is concerned, the basins with low potential ecological loss (1~2) are mainly distributed in the southwest of Dali, such as Yinjiang basin in Yongping and Xihe basin in Weishan (Figure 5b). These basins are limited by topography, with scattered cultivated land, relatively more broad-leaved forests, and relatively poor water conservation services and food supply services. The basin areas with potential ecological loss grade 3 are concentrated in the northwest of Dali Prefecture, such as Lancang River Basin in Yunlong and Shunbi River Basin in Yangbi. The topography of these basins fluctuates greatly, and the soil and water conservation of Lancang River, Nujianggao and Zhongshan Valley involved is obviously higher than that of Cangshan Erhai Plateau lake basin in the east. At the same time, the area ratio of15 ~ 25 slope belt is large, and the slope and hydrothermal conditions are suitable for the growth of various vegetation, which is conducive to the play of the advantages of net primary production and soil and water conservation services. The basins with high potential ecological losses (Grade 4~5) are concentrated in the northeast of Dali Prefecture, such as Sangyuan River Basin in Binchuan County, Zhonghe River Basin and Luolou River Basin in Heqing County. These basins are relatively flat, with vast arable land, dense distribution and good hydrothermal conditions. They are the main grain supply areas in Dali, with many mixed forests, high efficiency and large water conservation, so they have strong food supply services and water conservation services (Figure 5c).
4.4 Ecological risk of landslide disaster
Based on landslide disaster, ecological fragility and potential ecological loss, the ecological risk of landslide disaster in Dali basin scale is obtained by equal weight multiplication, and it is divided into three grades according to natural breakpoint method. At the same time, for the layers of danger, vulnerability and potential loss, 1~3 is merged into a low-value area, and 4~5 is merged into a high-value area, thus obtaining 8 types of ecological risk composition (Figure 6). For example, "high risk-low vulnerability-high loss" means that a watershed belongs to high-value danger zone, low vulnerability zone and high potential ecological loss zone.
Graphic options
Fig. 6 Ecological Risk Grade and Structure of Landslide Disaster in Dali Basin Fig. 6 Ecological Risk Grade and Structure of Landslide Disaster in Dali Basin
From the overall state, the spatial distribution of low-medium-high ecological risk of landslide disaster has a certain circle structure from the outside to the inside, and the basin area and the number of basins of each risk level decrease in turn (Figure 6). Among them, the low-risk areas are mostly distributed around Dali Prefecture, mainly concentrated in Nujiang River Basin in the southwest of Yunlong County, Heihuijiang River Basin in Eryuan County and Qingshuihe River Basin in the northeast of Xiangyun County. Medium-risk areas gradually gather in the center of Dali Prefecture in spatial distribution, mainly located in the Lancang River Basin in the southwest of Yunlong County, the Yinjiang River Basin in the middle of Yongping, the Luolou River Basin in Heqing County and the Xihe River Basin in Weishan County. High-risk areas are mainly concentrated in Sangyuan River Basin in Binchuan County and Yupaojiang River Basin in Xiangyun County.
By comparing the structural composition of each level of risk and the corresponding number of watersheds (Figure 6), we can know that high-risk areas include five risk components, with three high types of "high risk-high vulnerability-high loss" and two high types of "low risk-high vulnerability-high loss" as the main ones. The low-risk area consists of seven risks, including three types of low risk-low vulnerability-low loss and two types of low risk-low vulnerability-high loss, high risk-low vulnerability-low loss and low risk-high vulnerability-low loss. The medium risk area includes all seven types except "high risk-high vulnerability-high loss", and the quantity distribution of each type is relatively balanced.
5 discussion
5. 1 Verification of landslide risk assessment results
6 conclusion
In this paper, taking Dali Bai Autonomous Prefecture as an example, based on the three-dimensional risk assessment framework of "risk = danger × vulnerability× loss", the information model is used to assess the risk of landslide disaster, and the ecological vulnerability is characterized based on the landscape pattern index. The ecosystem service is included in the quantitative characterization of risk loss, and the ecological risk of basin geological disasters is distinguished through comprehensive measurement. Based on the leading risk factors, the ecological risk prevention zoning and risk prevention strategies were discussed. The results show that: ① The "dominant" conditions that are easy to induce landslide disasters in Dali are: the altitude is lower than 1800 m, the slope is15 ~ 25, the NDVI is less than 0.3 1, the lithology is mudstone, shale, loose sandstone, coal seam, volcanic debris and phyllite, and the nuclear density at the earthquake point is greater than 0. The annual average rainfall is1100 ~150 mm, within 1000 m from the fault zone, within 500 m from rivers and highways, and within 100 m from the town. Dali prefecture is generally in the middle and high-grade landslide disaster, with low northwest and high southeast; 36.3 1% of grade 4~5 high-risk areas are distributed in Weishan, Nanjian and Xiangyun County, and 54.6 1% of grade 2 low-risk areas are concentrated in Yunlong, Eryuan and Jianchuan. (2) The high-value areas of ecological vulnerability 4~5 are mainly concentrated in the south of Red River Basin, the southeast of Jinsha River Basin and the middle of Lancang River Basin. The low-water basin with ecological loss accounts for 43.23% of the whole basin, mainly distributed in the southwest of Dali Prefecture, and the water conservation and food supply services are relatively poor. The middle watershed area of ecological loss accounts for 28.9 1%, which is concentrated in the northwest of the state, and its advantages in net primary production and soil and water conservation services are obvious. The basins with high ecological losses are concentrated in the northeast, which has strong food supply and water conservation services. ③ The ecological risk of landslide disaster presents a low-medium-high ring structure distribution, and the basin area and quantity of each risk level decrease in turn; 367 small watershed has eight risk components, such as "low risk-low vulnerability-low loss" and "high risk-low vulnerability-high loss". Based on three risk levels and eight risk structures, Dali Prefecture is finally divided into four risk prevention types, namely, disaster prevention monitoring and early warning area, ecological protection and restoration area, disaster prevention protection area and natural adaptation control area.
Based on the two-dimensional framework of classical probability-loss model, this paper further emphasizes the exposure response of risk receptors to risk sources, and constructs a three-dimensional framework of "danger-vulnerability-loss" for ecological risk assessment, which provides some theoretical support for the enrichment of information in the process of quantitative research of ecological risk. In terms of research methods, vulnerability is expressed by landscape pattern, and potential loss is expressed by spatial field accounting of ecosystem services, which explains the mutual feedback relationship between ecosystem pattern and process, and is conducive to understanding the state changes related to the resilience of risk recipients under disaster stress at the mechanism level. In the evaluation unit, the watershed, the natural boundary of the surface, is taken as the boundary of the evaluation unit, which ensures the integrity of the structure and process of natural elements in the unit and the spatial heterogeneity of the natural environment between units, avoids the separation of the natural geographical connection of the surface, and is conducive to the overall grasp and comprehensive analysis of the risk pattern.
However, there are still some shortcomings in this study, especially the uncertainty analysis of risk assessment. It is necessary to further study the inevitable uncertainties in the ecological risk assessment of landslide geological disasters, such as the incompleteness of information and data, the diversity of loss types, the universality of risk sources and losses, and some random interference. Aiming at the source of uncertainty, Bayesian network simulation [46] and Monte Carlo simulation [47] are used to carry out uncertainty analysis and sensitivity analysis to help decision makers put forward more scientific and effective risk management countermeasures according to the uncertainty of evaluation results.