Developing Landslide Hazard Scenario of Kashmir Himalaya from the Historical Events

Kashmir Himalaya being a rugged and tectonically active zone has complex, unstable geology along with steep slopes, creating a favorable environment for landslide hazards, especially along the National Highway (NH-44) that connects the Vale of Kashmir with the rest of India. The historical landslide database for the whole country has not yet been developed and the data provided by various government organizations are often very limited because most of the time local and small-scale landslide events do not get recorded, thus, leading to misinterpretations. The present study focuses on retrieving the information on landslide events and their impacts to develop a comprehensive database for the period from 1990 to 2020 in Jammu and Kashmir, emphasizing Jammu-Srinagar National Highway (NH-44). A hotspot analysis tool (Getis-ord-Gi* algorithm) was used to understand the spatial distribution and concentration of the events throughout the region. The annual and seasonal analysis of the 739 landslide events reported in the valley for the selected period suggests an increasing trend causing 1000 fatalities and 267 injuries. The ndings show that out of 20 districts, 16 are relatively more exposed to landslides and the socio-impact induced by landslides was found more along the NH-44 with 303 landslide occurrences reported in 260 days in the past three decades having a high intensity of damage and loss. The results of this study are expected to be of potential use for developing a Landslide Early Warning System (LEWS) and for mitigating the impacts of landslides in the Kashmir Himalaya.


Introduction
UNDRR has de ned landslide as the downhill movement of a mass of rock, debris, or earth down a slope under the direct in uence of gravity, ranging from rapidly moving catastrophic rock avalanches and debris ows in mountainous regions that can destroy property and take lives suddenly and unexpectedly, to more slowly moving earth slides like creeps that may cause damage gradually. The elements that make an area prone to landslides generally include bedrock geology, geomorphology, land use and land cover, type of soil, and precipitation (Chingkhei et al. 2013). Landslides cause loss of human lives, damage to property, road infrastructure, and may block the streams and rivers as well as impacting the water quality (Righini et  McCalpin 1984). Landslide inventory maps also help in generating the hazard scenario of any region (Antonini et al. 1993; Remondo et al. 2008). Based on the accessible resources, landslide inventories are prepared in several ways using different techniques which incorporate eld survey, visual interpretation of aerial photographs, and collected data from historical archives and by using remote sensing techniques season. Among the Asian countries, India is highly exposed to landslides (Dikshit et al. 2020; Bhandari, 2006) and the area most vulnerable to fatal landslides is found to be the west and the north-west Himalayas, followed by east and north-east India and South-India indicated by a study (Prakash 2011).
According to National institute of Disaster Management in the year 2011, an estimate suggested that India suffers a monetary loss of Rs 150-200 crore every year from landslide and by the Geological Survey of India, 420,000 sq km, or 12.6% of the total land area, is landslide-prone, which spreads over more than 65,000 villages in hilly/mountainous regions of the country (NDMA). Out of the total landslide-prone area of India (12.6%), North-western Himalayas constitute 33%, which grasps the attention of researchers to study this part of the Himalayas. In a data-scarce environment like India, there is still a lack of comprehensive documentation of historical landslide events. The available data has huge time gaps and are at different scales, which create a barrier in the reconstruction of past scenario of the hazard prone areas ( Van et al. 2013

Study Area
The Himalayas are the youngest mountain range in the world formed due to the subduction of the Indian plate under the Eurasian plate that started during the Eocene epoch (Gansser 1964;Bhat 1987

Material And Methods
Landslide Database Inventory of Kashmir Himalayas has been compiled for the time period of 1990-2020 with special reference to Jammu-Srinagar national highway (NH-1A). The data for generating the landslide Database has been collected from multiple sources mainly: National newspapers (Hindustan The methodology adopted in this paper set off with the collection of historical events for which the newspaper archive of the Central Library of the University of Kashmir was utilized (Taylor et al. 2015).
The newspapers available from the 1990s were selected for the study and the rest of the data used was  (Sultana 2020). Along with the Landslide episodes, some major fatal avalanche episodes that may have triggered landslides were also documented in the database. Landslide inventory maps were generated using Arc-GIS software. Annual, decadal, and seasonal trends and frequency of the landslide events were analysed. District-wise landslide maps were prepared with the help of GIS techniques to nd the districts most susceptible to fatalities and injuries caused by landslides and landslides associated with avalanches. Also, a catalogue of major fatal events of landslide/avalanches was prepared to help in the management of landslide disaster risk reduction. A Hotspot analysis tool (Getis-ord-Gi* algorithm) was used to map the spatial distribution and concentration of events, fatalities, and injuries by the Jammu-Srinagar National Highway NH-44 being highly prone to landslides was studied individually. Along the national highway NH-44, dozens of places get hit by multiple landslides on a single date which is not reported individually. Due to the lack of reported data, the estimation of the no. of landslides has been calculated by multiplying the days on which land sliding was reported (LD) with the number of places hit (PH) at a particular date. The socio-economic impact of landslides along the highway has been shown by the intensity of the damage caused throughout the time period 1990-2020 based on the obtained data.

Measurement of damage intensity
The socio-economic impact in terms of monetary losses caused by landslides was not available and to calculate the intensity of damage over the time period 1990-2020, a general scale of four classes was devised for measuring the impact in the area of interest prepared in accordance with the damage and loss database given as below (Table 3).

Results
There is an asymmetrical distribution of landslide events along the Kashmir Himalayas. From the landslide database prepared for the Kashmir Himalayas during 1990-2020, it was found that the entire region is prone to landslides that are largely distributed along the major transport routes passing through Greater Himalayas and PirPanjal ranges including; NH-44 (Jammu Srinagar national highway), Mughal Road, NH-1D (Srinagar-Sonmarg-Gumri Road, NH-1B (Sinthan Top-Kishtwar-Batote Road), Bandipora-Gurez Road, Baramulla-Uri Road (NH-1A) and Kupwara-Machhil Road (Fig. 2a). The study area lies within administrative boundaries of the Union Territory Jammu and Kashmir comprising 20 districts. The geographical distribution of total landslide events and of major catastrophic fatal events is shown in Figs. 2(a) and (b). The Landslide Database of Kashmir Himalayas (LDKH) identi es about 739 landslide events reported in 506 days at the regional scale ( Table 4). The most affected local areas were the Banihal-Ramban stretch, Baltal route (Ganderbal), Sonmarg-Gumri route (Ganderbal), Uri (Baramulla), Dawar-Gurez (Bandipora), and Kishtwar route outlined in Table 4. Out of the 739 landslide events, 180 fatal events were identi ed from 1990-2020 that were discussed in detail in the fatal landslide Catalogue.
Preparing the inventory for fatal landslides was a challenge because most of the fatal landslides were associated with avalanches and documented in landslide disaster reports, therefore making it di cult to isolate major avalanche events from landslides (Prakash 2011 The total deaths caused by landslides and landslides associated with avalanches counted 1000 (including Catastrophic events) and the total number of reported injuries were 267 (Table 4). Fig. 3 represents the hotspot analysis of historical landslide events, fatal landslides and fatal landslide/avalanche events (L/A) for the study period 1990-2020 in the Kashmir Himalayas. The results identi ed three historical landslide hotspots on the main transport routes as Srinagar-Jammu highway (Banihal to Udhampur), Srinagar-Sonamarg-Gumri road and Bandipora-Gurez road (Fig. 3). The concentration of the event occurrences was found low in the valley oor and dense along the transport routes. Considering the LDKH, the Landslide fatality hotspot lies on the Jammu-Srinagar Highway from Banihal to Udhampur and the landslide/ Avalanche fatality hotspot lies in the Gurez Valley (Bandipora).
The district-wise spatial distribution of fatal landslide events and the frequency of deaths and injuries caused by them has been shown in Fig. 4. Fig. 4(a) shows that the districts with the most frequently occurring fatal landslide events from 1990-2020 were Ramban with 24 fatal events and Baramulla with 15 fatal events, followed by district Kishtwar (14), Doda (13), Reasi (12), Ganderbal (9) and the districts with an average of 5 to 9 events were Poonch, Kupwara and Anantnag. The districts with zero occurrences of fatal Landslide events were Jammu, Kathua and Shopian. From the Analyses of the database the highest frequency of deaths caused by landslides from 1990-2020 were reported in District Ramban (44), Kishtwar (38) followed by District Doda (32), Reasi (27), Poonch (29), Baramulla (24), Ganderbal (19), Rajouri (18) and districts with Minimum fatalities were Udhampur, Samba, Pulwama and Srinagar (Fig. 4c).The districts with the highest number of injuries were identi ed as Ramban with 57 reported injuries and Reasi with 33 reported injuries followed by Kishtwar (21), Ganderbal (17), Baramulla (16) and Doda (14) (Fig. 4d).
The major catastrophic events were separated that could have triggered the landslides, where the death toll crossed 30 to 200 deaths /per event, were stroked in the districts Kulgam and Bandipora. Figs. 4(b), (e) and (f) show the frequency of landslide/avalanche (L/A) events (landslide associated with snow avalanches), fatalities and injuries caused by them from 1990-2020. The results show that the highest frequency of fatal L/A events was seen in districts Kupwara with 14 fatal L/A events, Bandipora (11) and Kulgam (6) followed by Aanatnaq (5), Ganderbal (4) and the district with the minimum occurrence of fatal events has been found in Ramban, Reasi and Doda (Fig. 4b). It has been noticed that the districts with the highest fatalities caused by L/A events were Bandipora with 67 deaths, Kupwara (53) and Budgam (40), followed by Kulgam (31), Baramulla (25) and Reasi (17) shown in Figs. 4(e) and (f). The South -West of the Kashmir Himalayas was found to be less prone to avalanches and the landslides associated with them. Fig. 4(f) represents the spatial distribution of the injuries caused by L/A which was found to be highest in districts Anantnag, Budgam, Bandipora and Baramulla.  found, one with 99% con dence and the other two with 95% and 90% con dence level for the historical landslides lying between Nashri-Ramsu.  (43). From the analysis of the seasonal variation in the frequency of the landslides from 1990-2020, the maximum number of landslides were reported in the winter season (139) and monsoon (83) followed by pre-monsoon (69) and the lowest was found in post-monsoon (12).

Socio-economic impact
To examine the socio-economic impact of landslides along NH-44 in the Kashmir Himalayas, the intensity of the damage has been measured from 1990-2020 through a scale prepared using the data gathered from different sources given in Table 3.The socio-economic impact caused by landslides in the region is discussed in detail in Table 5 and from which it is evident that the Jammu-Srinagar highway has proven disastrous to the people of the region. It has been noticed that from 2007 onwards the intensity of damage has been very high from 1990-2006 the damage intensity ranges from low to high (Fig. 9 b) and also, the number of causalities has increased from 2007 onwards.
From the database (Table 5) it is manifested that the highway remains closed for most of the winter due to landslides which hike the prices of the necessary commodities in the basin leading to short-term in ation. The passengers, supply trucks, oil tankers, and tourists get stranded on the highway for dozens of days in the winter season with no availability of food and water. The stranded passengers sometimes travel kilometers on foot to reach their homes. During monsoon season lacs of pilgrims who visit the Amarnath temple in the Kashmir valley frequently get stuck on the highway. The majority of the time Yatra is suspended for many days due to landslide incidents. The landslide fatality database shows that many pilgrims have lost their lives in these past years. Besides government of India started the road widening project of NH-44 in the year 2011 which has exposed new landslide-prone areas making the situation more depressing (Pandey et al. 2014). Due to the increased frequency of landslides, the cost for the maintenance of the highway has also increased.  The probable reason for the high susceptibility of landslides in the Himalayas is because the region lies in seismic zone 5 where we can experience earthquakes often resultin in huge mass movements (Rashid et al. 2017).The south-facing slopes of Himalaya that are largely exposed to Freeze-thaw conditions and have a lack of vegetation cover are more prone to landslides (Ambraseys and Bilham 2012 1990-2020. The seasonal distribution of landslides reveals that the winter season has recorded the maximum number of landslides which is about 33.2% of the total events followed by the monsoon (32.7%) and pre-monsoon (29.5%). The post-monsoon season depicted the least no. of landslide incidents in the region, only 5% of the total landslides. The district-wise spatio-temporal pattern of the landslides was studied, out of 20 districts, 16 districts were found to be more vulnerable to landslides.
The study speci cally identi ed 303 landslides reported in 260 days in the Jammu-Srinagar Highway from Udhampur to Banihal and the total estimated landslides for the highway were 1482. The spatial distribution of the landslide incidents analyses that section Marroog-Ramsu has the highest concentration of events from 1990-2020. An increasing trend of landslides was found with the highest frequency in the winter season accounting for 45.8% of total landslides, and monsoon season (29.3%) followed by pre-monsoon(22.7%) and the lowest frequency of landslides was found in the post-monsoon season (3.9%). The socio-economic damage caused by landslides annually is shown in table (5)       showing Damage intensity annually caused by landslides along NH-44 from 1990-2020