Characterization Of Landslide Hazard In The Sub Himalayan Zone Of Nepal

Abstract

Landslides are a typical geological process in the Himalayan Mountains, causing yearly loss of life and property during the Monsoon period. Active tectonism, seismic activity, monsoon rainfall, complicated geology, and steep slopes are crucial markers of the Himalayan landslide process. The Siwalik is the world's youngest mountain range, comprising fragile sedimentary rocks. Landslides are a common natural disaster in the Sub Himalayan zone (Siwalik) during the Monsoon period, resulting in the loss of lives and property. Landslide mitigation and prevention are a significant challenge in the Siwalik region. The characterization is a new issue for this region to explore and comprehend the natural process of landslides. This study aims to characterize the landslide by using geological, topographical, hydrological, and geotechnical attributes along with landslide classification and susceptibility modeling in the Babai River watershed of the Siwalik region.

The procedure began with the spatial and temporal landslide inventory mapping by using Google Earth, Landsat, and Sentinel imageries. The spatial inventory map included all landslides in the study area, whereas the temporal inventory map included new landslides annually from 2010 to 2021. The size, length, area, and number of landslides in different geological units were investigated from the landslide inventory map. In addition, the mechanism of landslide beginning in geological units was studied. Total yearly and monsoon rainfall amounts, the geographical distribution of rainfall, and the distance to the drainage map were used as hydrological characteristics. Data from seven stations from 2010 to 2021 were used to create the rainfall distribution map. The slope, aspect, curvature, topographical wetness index, and “T” factors were used for topographical characterization. The shear behavior, phase relation, and index properties of 120 landslide soils were obtained for geotechnical characterization. A size, activity, and mechanism-based classification of the landslide were proposed. For the susceptibility analysis, 14 landslide-causing factors were chosen. Similarly, four bivariate models were used for landslide susceptibility analysis: the weight of evidence, frequency ratio, information value technique, and Shannon's entropy. The silt/clay ratio in Siwalik's soil is one of the leading causes of landslides in saturated conditions. The soils of the Middle Siwalik are stiff, whereas those of the Lower and Upper Siwalik are soft. Lower and Upper Siwalik soils are similarly poorly graded. The average shear strength in the Lower and Upper Siwalik is also lower than in the Middle Siwalik. Minor and shallow landslides characterize the Lower and Upper Siwalik. Large-scale landslides are prevalent in the Middle Siwalik. The "T" factor and landslides have a positive association. There is a substantial positive relationship between total annual rainfall and yearly landslides. Similarly, hill slope and landslide incidents are strongly correlated. A size-based classification was created based on the landslide area's logarithmic scale. Size-based classification is proposed in five sizes: very small, small, medium, large, and very large. Landslides with less than 100 m2 of surface area are considered very small, while those with more than 100,000 m2 are considered very large. For activity-based classification, five activity states are recommended: new, active, inactive, stable, and reactivated. It describes the intended activity state conditions. Based on the landslide mechanism, five types of landslides have been proposed: Type 1, Type 2, Type 3, Type 4, and Type 5.

The percentage of very high susceptibility (21%) is obtained from the information value method, whereas the low susceptibility percentage (8%) is obtained from the Shannon entropy model. Medium susceptibility is obtained higher from the information value, weight of evidence, and frequency ratio models. The result shows that the susceptibility distribution is more or less similar between frequency ratio, the weight of evidence, and the information value model; however, the Shannon entropy model has a slightly different result. The weight of the evidence model has a higher accuracy and prediction rate, according to the results. The WoE model has a success rate of 85% and a prediction rate of 79.9%. The accuracy and prediction rate of the remaining three models are comparable. The prediction rate hierarchy is WoE (79.9%), FR (75.3%), IVM (74.4%), and SE (73.2%). The success rates of four different models (WoE, FR, IVM, and SE) are 85%, 78.75%, 78.57%, and 77.2%, respectively. Every model's accuracy and prediction rate are greater than 70%, indicating that all models can be used to predict landslide susceptibility in Nepal's Siwalik Hills. हिमालय क्षेत्रमा पहिरो खस्नु एउटा निरन्तर भौगर्भिक प्रक्रिया हो । नेपालको सवहिमालय (सिवालिक) क्षेत्रमा वर्षायाममा ठूलो मात्रमा पहिरो जाने गरेको छ जसले गर्दा हरेक बर्ष मान्छेको जिउ धनको क्षति हुने गरेको छ । सक्रिय टेक्टोनिक्स, भूकम्पीय गतिविधि, मनसुनी वर्षा, कमजोर भौगर्भिक अबस्था र ठाडो भिर युक्त जमिनहरु हिमालय भूस्खलन र पहिरो जाने प्रक्रियाका महत्वपूर्ण अवयवहरु हुन् । यस अध्ययनले नेपालको पश्चिम क्षेत्रको सवहिमालय (सिवालिक) भित्र पर्ने बबई नदीको जलाधार क्षेत्रमा पहिरोको वर्गीकरण र संवेदनशीलता मोडलिङका साथै भौगर्भिक, जलविज्ञान, भूसतहिक र भूप्राविधिक विशेषताहरु प्रयोग गरी पहिरोको विशेषता निर्धारण गर्ने लक्ष्य राखेको छ । अध्ययनको सुरुवात पहिरोको सूची नक्साङ्कनबाट गरियो । जसमा गुगल अर्थ, ल्याण्ड स्याट र सेन्टिनल इमेजरीहरुको प्रयोगबाट इस्वी सम्बत २०१० देखि २०२१ सम्मका नयाँ तथा पुराना पहिरोहरुको स्पेश र टाइमका फरक नक्साहरु तयार गरियो । डिजिटल एलिभेसन मोडलको प्रयोगबाट पहिरो जाने सम्भावित फ्याक्टरहरुको नक्सा तयार गरियो । भौगर्भिक नक्सा तयार गरी तीनवटा फर्मेशनहरु क्रमश लोयर सिवालिक, मिडल सिवालिक र अपर सिवालिक साथै चट्टान नबनेका तटीय सेडिमेन्टहरुलाई quaternary deposit मा छुट्याइयो । अध्ययन क्षेत्रको पुरै भाग समेटिने गरी प्रत्येक फर्मेसनबाट १०,००० वर्ग मिटर भन्दा बढि क्षेत्रफल भएका ४० वटा पहिरोहरु छुट्याई स्थलगत अध्ययन गरी माटोको नमुना सङ्कलन गरियो । सङ्कलित नमुनाहरुलाई भूप्रयोगशालामा ल्याई विभिन्न भूप्राविधिक परीक्षणहरु गरियो । पहिरो संवेदनशील नक्साङ्कनका लागि १४ वटा सम्बन्धित कारकहरुको नक्सा तयार गरियो र पहिरो प्रकोपको अवस्था थाहा पाउन ४ वटा bivariate model हरु छनोट गरियो । यस अध्ययनको नतिजाबाट समग्र बार्षिक वर्षा र समग्र बार्षिक पहिरो सङ्ख्या बिच पर्याप्त सकारात्मक सम्बन्ध देखियो । पहाडको पन्ध्र डिग्रीदेखि पैंतालिस डिग्रीको स्लोप बढी संवेदनशील देखिन्छ । दक्षिण aspect मा बढी पहिरो जाने गरेको भेटियो भने concave सतह भएको भिरालो जमिनमा बढी सङ्ख्यामा पहिरो जाने गरेको पाइयो । माटोमा पाइने सिल्ट/क्ले अनुपात वर्षाको पानीले पहिरो गराउने प्रमुख कारण रहेको पाइयो । Lower र Upper सिवालिकका माटोहरु नरम र कम घनत्वका भेटिए भने Middle सिवालिकका माटोहरु कडा र बढी घनत्वका भेटिए । समग्रमा Middle सिवालिकमा औसत shear strength parameter बढी देखियो जसले गर्दा चट्टानी पहिरो बढी तर माटोको कारणले जाने पहिरो न्यून सङ्ख्यामा देखियो । पहिरो जाने सम्यन्त्रका आधारमा सिवालिक क्षेत्रका पहिरोहरुलाई पाँच समूहमा बर्गीकरण गरियो । त्यसैगरी आकार र गतिशीलताका आधारमा पनि पाँच समूहमा बर्गीकरण गरियो । अध्ययन क्षेत्रमा चारै ओटा मोडलका पहिरो संवेदनशील नक्साहरु तयार गरियो जसमा weight of evidence मोडलले frequency ratio, information value र entropy मोडल भन्दा उत्तम नतिजा दिएको छ भने चारैओटा मोडलहरु पहिरो जोखिम अध्ययनको लागी प्रयोग गर्न सकिने देखिएको छ । सबै मोडेलहरुले पहिरोको जोखिम बीस देखि तीस प्रतिशतको हाराहारीमा देखाएको छ । यो अध्ययनले नेपालको सिवालिक क्षेत्रको पहिरोको बहुआयामिक तवरले विशेषीकरण गरेको छ र उक्त क्षेत्रका पहिरो अध्ययन र अनुसन्धान तथा नियन्त्रणमा महत्वपूर्ण भूमिका खेल्न सक्ने छ ।