2018
Information Value Model based Landslide Susceptibility Mapping at Phuentsholing, Bhutan
PASANG, Sangey a Petr KUBÍČEKZákladní údaje
Originální název
Information Value Model based Landslide Susceptibility Mapping at Phuentsholing, Bhutan
Název česky
Information Value Model based Landslide Susceptibility Mapping at Phuentsholing, Bhutan
Autoři
PASANG, Sangey a Petr KUBÍČEK
Vydání
Lund, Sweden, Association of Geographic Information Laboratories in Europe (AGILE) 2018, od s. 1-7, 7 s. 2018
Nakladatel
Association of Geographic Information Laboratories in Europe (AGILE)
Další údaje
Jazyk
angličtina
Typ výsledku
Stať ve sborníku
Obor
10500 1.5. Earth and related environmental sciences
Stát vydavatele
Švédsko
Utajení
není předmětem státního či obchodního tajemství
Forma vydání
elektronická verze "online"
Označené pro přenos do RIV
Ne
Organizační jednotka
Přírodovědecká fakulta
Klíčová slova česky
Landslide susceptibility map, Information value model, Geographic information system, Area under curve.
Klíčová slova anglicky
Landslide susceptibility map, Information value model, Geographic information system, Area under curve.
Příznaky
Mezinárodní význam, Recenzováno
Změněno: 1. 12. 2020 12:16, Sangey Pasang, Ph.D.
V originále
In the current study, statistical method of information value and geographic information system (GIS) were applied to develop Landslide Susceptibility Map (LSM) of Phuentsholing region, Bhutan. A total of 161 landslides, covering an area of 2.92 square kilometres were identified and 20% was randomly extracted for validation. Various factors causing landslide such as slope, aspect, elevation, proximity to road, drainage and fault, lithology, land use and normalised difference vegetation index (NDVI) were analysed to determine the contribution of each factors to the occurrence of a landslide. To evaluate the performance of the information value model in determining the LSM, overlay method and the Area under curve (AUC) of the receiver operating characteristic (ROC) were performed on the training and validation samples. The region was categorised mostly under high and moderate susceptibility, with land use, vegetation and elevation identified as most important contributing factors to landslide occurrences. The model has an AUC accuracy of 83.4% success rate and 83.5% prediction rate, with 77.5% of the validation samples lies under very high and high landslide susceptibility area when overlaid on the LSM
Česky
In the current study, statistical method of information value and geographic information system (GIS) were applied to develop Landslide Susceptibility Map (LSM) of Phuentsholing region, Bhutan. A total of 161 landslides, covering an area of 2.92 square kilometres were identified and 20% was randomly extracted for validation. Various factors causing landslide such as slope, aspect, elevation, proximity to road, drainage and fault, lithology, land use and normalised difference vegetation index (NDVI) were analysed to determine the contribution of each factors to the occurrence of a landslide. To evaluate the performance of the information value model in determining the LSM, overlay method and the Area under curve (AUC) of the receiver operating characteristic (ROC) were performed on the training and validation samples. The region was categorised mostly under high and moderate susceptibility, with land use, vegetation and elevation identified as most important contributing factors to landslide occurrences. The model has an AUC accuracy of 83.4% success rate and 83.5% prediction rate, with 77.5% of the validation samples lies under very high and high landslide susceptibility area when overlaid on the LSM
