Introduction
Numerous forms of natural dangers and related tragedies containing earthquakes, volcanic eruptions, tsunamis, cloud burst, floods, soil erosion etc. happening the world and amongst such troubling natural hazards landslides are the awful types of greatest recurrent occurrences all everywhere the world. Every year, landslides have affected huge damages of life and stuff, concluded the damages of forests, fruitful cultivated land, habitation area, and network communication in addition to tourist adverts. Additionally, alteration of the earth surface is also responsible for devastating landslides. Iran has confronted numerous categories of natural threats and disasters, for example severe soil erosion concluded gully expansion, vulgar floods, and disturbing landslides. So, because of the numerous occurrences of landslides and huge financial damages have develop national disasters of Iran. The landslide event in Iran has caused about 500 billion financial damages. Landslide susceptibility assessment be able to assistance the planners for final management of environmental squalor and natural resources from delicate damages and eventually development of economic action of this watershed area. The planned methods use both the capability thoughts and ground fact at the like time. This could be taken as a brand-new methodology toward landslide zoning difficulties. The goal of the study is to zonate the hazard of landslide occurrence using maximum entropy (ME) algorithm and compare the effectiveness of this method in locating the sensitivity of landslide occurrence in Karganeh Watershed, Lorestan Province.
Materials and Methods
Karganeh Watershed is located between 33° 25′ 12″ to 33° 37′ 12″ latitude and 48° 23′ 59″ to 48° 44′ 24″ longitude, occupying about 294.2 sq km in the Lorestan Province, west of Iran. This watershed is one of the main sub basins of Karkheh River. One of the most important phases of landslide susceptibility assessment is to identify and prepare a distribution map of current landslides in the watershed. This map was prepared via assembly the information associated with landslides or via analyzing the data from remote sensing and GIS techniques. On behalf of this goal, the distribution map layer of landslides in the watershed was prepared and separated into two sets for model training (70%) and model validation (30%) randomly. Also, 16 causes disturbing the happening of landslides in this watershed were selected permitting to the review of sources and the usage of principal component analysis (PCA), Tolerance and VIF tests. Digital layers of effective factors in geographic information system were equipped. In the next step, the landslide risk map was equipped based on the maximum entropy machine (ME) method. So that evaluate the accurateness of the modeling and compare the efficiency of the method, the index of the area under the virtual performance recognition curve (ROC) was used. Established on the fallouts of the maximum likelihood diagram, geological, land use and slope are the best significant factors inducing the event of landslides in Karganeh Watershed.
Results and Discussion
Landslide inventory map indicated that there are 95 scattered landslides in the Karganeh Watershed. Exaggerated total area through landslide is 635 ha (2.23% of the watershed area). Based on the results of the PCA index, the KMO coefficient was calculated as 0.61, which confirms the necessary correlation between the input variables to perform principal component analysis. Among the 19 components as the number of variables investigated in landslide risk assessment, considering the eigenvalue higher than one, the number of the first two main components was investigated. The results showed that these two main components express nearly 67% of the changes. After investigation, slope, slope direction, elevation classes, geology, distance from the river, distance from the road, distance from the fault, river power index (SPI), topographic moisture index (TWI) and slope length index (LS), topographic position index (TPI), topographic roughness index (TRI) and vector roughness measurement index (VRM), land use, distance from the village, and rainfall were selected as the most effective factors of landslide occurrence in the Karganeh Watershed. According to the Kappa index diagram, geological indicators, land use, slope, topographic roughness index (TRI), slope length and slope direction are the most important influencing parameters. The area under the curve (AUC) based on the relative performance detection curve indicates good accuracy (AUC=0.787) in the validation stage. Permitting to the results of the maximum entropy method, about 28% of the Karganeh Watershed is in the high and very high hazard class of landslide happening.
Conclusion
In this study, it was tried to use all effective factors in order to evaluate landslide susceptibility in Karganeh Watershed. Principal component analysis (PCA), tolerance and VIF were used to determine the relationship between the factors influencing the occurrence of landslides and to determine the most effective factors. It is the carrying capacity of the waterway. It actually determines the effect of topography on erosion. The longer the slope is, the higher the sediment carrying capacity is, and the condition of landslides on the side of the waterway increases. An increase in the topographic roughness index (TRI) indicates uplift and nontectonic activity. Changing the rangeland to rain fed farming and road construction is completed severely in Karganeh Watershed throughout recent years and led to awarding great role of human factors on landslide in comparing other factors. Based on the results obtained from this model, 50.4% of the area of the basin is in the very low and low sensitivity class, 21.6% is in the medium sensitivity class, and 28% is in the high and very high sensitivity regional level. An increase in this index leads to more surface roughness and an increase in slope, which makes landslides more prone to occur.The implementation of landslide management programs based on the results of this research on the Karganeh Watershed can explain the difficulties of domain instability. |
