Indicator-Based Assessment of Resilience and Vulnerability in the Indian Himalayan Region: A Case Study on Socio-Economy under Di ﬀ erent Scenarios

: The Indian Himalayan region is vulnerable to climate change because of its geospatial fragility. The present study gives a framework for the analysis of household and village-level resilience and vulnerability in the Bhagirathi Basin of Indian Western Himalayan region under di ﬀ erent climate change scenarios. Villages were selected depending on di ﬀ erent biophysical criteria to have a good representation of the study area. Household-level survey using the household economy approach was done in 646 households of 30 villages to collect information on indicators of natural, physical, ﬁnancial and human capital assets and scores were generated for each category. A cumulative resilience score was obtained for each household and village. Future climate projections on mean annual temperature were also accessed under Representative Concentration Pathway (RCP) 4.5 to estimate the change in mean temperature of the studied villages and probable change in agricultural production. The result shows that most of the villages of Tehri Garhwal are clustered in vulnerable classes in comparison to Uttarkashi villages and vulnerability scores of 11 and 8 villages changed under climate shock and future agricultural production change scenarios, respectively. The study has manifold implications on further research and policy implementation under socioeconomic vulnerability in the Himalayan region. The index-based approach to assess to change The present study has explored the analytical utility of using index-based assessment of adaptive and provides an assessment tool that can be on a local and assess and village-level The methodology can also be adopted to compare and contrast adaptive capacity in larger spatial scale and di ﬀ erent settings. The study identiﬁes both the and the factors responsible for the resilience and that can be used and integrated into microplanning and macro-policy development for better resource allocation to the vulnerable community as part of the change and adaptation planning.


Introduction
Humans and the environment have been closely associated with each other since antiquity. In this symbiotic relationship, humans have tried to adjust themselves with the environment initially, but subsequently molded the environment according to their needs. This has resulted in imbalances in the environment and the various ecological systems therein, thereby inviting several manmade environmental insecurities. The continued global rise in greenhouse gas emissions projected in most countries in the twentieth century ensures the unavoidable circumstances of climate change [1,2]. Climate change can, and is, accentuating the occurrence of extreme climate events throughout the globe [3,4]. It has mixed effects on the local changes in cropping pattern, availability of multiple ecosystems services like water supplies, vegetation and forest, biodiversity and system health, modifications in the economy, social and political system [5]. Global climate change presents a challenge to the future livelihood and existence of human beings, especially for those who are currently living in underdeveloped areas [6]. Climate change hazards also will increase unprecedentedly to increase the vulnerability of all the people depending on marginalized resources as livelihood and health high. Whereas on the other hand inductive methods requires a large set of indicators, and thus the subjectivity and accuracy became less [50][51][52][53]. Climate change vulnerability analysis majorly has two component the innate resilience of the society or household and the shock or stress imparted from climate change [6,29,54]. The major objective of the study is to establish an indicator-based approach to assess the adaptive capacity and vulnerability of selected households and villages that can be replicated in a larger spatial scale for identifying climate sensitive areas for effective adaptation and mitigation planning for better sustainability. In this study, an attempt was made to assess the village and household-level vulnerability in the Bhagirathi basin of the Indian Himalayan region (IHR). Indicators for vulnerability assessment were selected on the basis of sustainable livelihood framework and IPCC (Intergovernmental Panel for Climate Change) 2014 guideline from both climate shock and adaptive capacity components. Both Secondary information from census data of India and primary household and village-level survey data on selected indicators were used to assess the adaptive capacity and vulnerability of villages in the IHR.

Study Area
The study was carried out in the Bhagirathi Basin of Uttarakhand, (Uttarkashi and Tehri Garhwal District) India ( Figure 1). Uttarakhand is a state of the Northern part of India with an area of 53,483 km 2 . The terrain of the state is mountainous with dry soils and the climate and ecology vary greatly with elevation and slope. Bhagirathi River Basin (catchments) represents different biogeographic provinces (1B and 2B) [55] and physical as well as anthropogenic gradients and it has seven different subbasins, i.e., Bhagirathi I, II, III and IV, Asiganga, Balganga and Bhilangana covering approximately 10,000 km 2 area and different land cover and land use patterns including human habitation, agricultural land, large and small water reservoirs, rivers, streams and wetlands, subtropical and temperate forests, alpine rangelands, glacial moraines, permafrost areas and trans-Himalayan cold deserts. The study area encompasses a wide range of elevation zones starting from 500 m at subtropical forests to 5000 m at trans-Himalayan cold deserts and accordingly represents a mosaic of several microclimatic regimes. Uttarkashi and Tehri-Garhwal both the districts are situated in the Garhwal region of the state of Uttarakhand. Both the districts are traversed by tributaries of mighty river Ganga and the economy of the region is largely depending on agriculture. Uttarkashi district has both the sources of river Ganga and Yamuna. The district headquarter Uttarkashi city lies in the main route of Gangotri and Yamunotri and attracts millions of Hindu Pilgrims every year. The district has a population density of 41 km -2 and decadal growth rate from 2001 to 2011 was 11.75% ("district census 2011", http://www.census2011.co.in/district.php). Tehri Garhwal, having administrative headquarter is at New Tehri, is one of the most highlighted districts of Uttarakhand because of the Tehri dam.
The villages of the Bhagirathi basin differ both in biophysical and socioecological characteristics and so does the livelihood of the people. In the high-altitude areas, horticulture and pastoralism prevail, whereas in low altitude areas, agriculture. The economy of the villages near the district headquarters (Uttarkashi and New Tehri, India) and the villages far away also differ accordingly and so does the livelihood patterns of the villagers. Within and between village variances are also quite high since from historical times different groups of people like Gorkhas, Gujjars, Pundir, Rajputs, etc. have invaded the Garhwal region and later became residents. Thus, along with Garhwali culture other cultural inheritance also preserved in the study area. The villages of the Bhagirathi basin differ both in biophysical and socioecological characteristics and so does the livelihood of the people. In the high-altitude areas, horticulture and pastoralism prevail, whereas in low altitude areas, agriculture. The economy of the villages near the district headquarters (Uttarkashi and New Tehri, India) and the villages far away also differ accordingly and so does the livelihood patterns of the villagers. Within and between village variances are also quite high since from historical times different groups of people like Gorkhas, Gujjars, Pundir, Rajputs, etc. have invaded the Garhwal region and later became residents. Thus, along with Garhwali culture other cultural inheritance also preserved in the study area.

Methodology
Village-level information was collected from the Census India data set available in the census India website (http://www.censusindia.gov.in/) based on countrywide census carried out in 2011. Information on the number of households, total population, presence of medical facility, educational institute, market, road connectivity, nearest town and other important facilities of 1096 villages within the study area were collected. Locations of 824 villages out of 1096 were validated using Google Earth Pro in the study area, as the rest of the villages were not found in the Google Earth Pro database. To examine the access to resources and how it varies to different villages remoteness index was calculated for each village on the basis of weighted score of the availability of basic facilities using Equation (1). The information on presence, absence and distance from the motorable road, hospital, health center, primary and secondary school, market, police station, post office, bank, nearest town and tourist spot were used for calculating the value of remoteness index.

Methodology
Village-level information was collected from the Census India data set available in the census India website (http://www.censusindia.gov.in/) based on countrywide census carried out in 2011. Information on the number of households, total population, presence of medical facility, educational institute, market, road connectivity, nearest town and other important facilities of 1096 villages within the study area were collected. Locations of 824 villages out of 1096 were validated using Google Earth Pro in the study area, as the rest of the villages were not found in the Google Earth Pro database. To examine the access to resources and how it varies to different villages remoteness index was calculated for each village on the basis of weighted score of the availability of basic facilities using Equation (1). The information on presence, absence and distance from the motorable road, hospital, health center, primary and secondary school, market, police station, post office, bank, nearest town and tourist spot were used for calculating the value of remoteness index.
where β is the weighted score of the basic amenities of the village and A is the score (presence absence or coded) of basic amenities of the village. A two-step algorithm was used to cluster the villages on the basis of the total geographical area of the village, population, the altitude of the village and remoteness index of the villages. Three clusters were formed on the basis of the 4 criteria given for analysis ( Figure 2). Cluster 1 comprises of 49 villages having high altitude, large geographic area, remote and less population. Cluster 2 comprises 11 villages having low altitude, easily accessible, moderate geographic area and high population. Cluster 3 again comprises of 49 villages having moderate elevation, small sized, moderately remote, sparsely populated. Climate change will increase the exposure of different villages to environmental hazards and will affect the agricultural sector more where irrigation facilities are limited. Conflict with wild animals and exposure to disasters will also increase due to climate shock. The villages having high dependency on agriculture, more prone to natural disasters and high wildlife conflict are more exposed to climate shock compare to the villages that lack those exposures. The village and household vulnerability also will be shaped by these factors. To include the factors associated with exposure to climatic stress in vulnerability analysis 102 villages from different clusters were selected for the primary data collection on the agricultural dependencies of the village, disaster proneness of the village and wildlife interaction.
On the basis of this information another clustering was done using a two-step algorithm. A total of 10 sub-clusters were formed, 5 from each Cluster 1 and 3, (Table S2 of supplementary). Due to the smaller number of villages in Cluster 2 (11 villages) it remained as a single cluster. Thirty villages were selected representing all the clusters for detail household-level survey using household economy approach ( Table 1). The household economy approach (HEA) is developed by Save the Children, UK and designed to get information based on livelihood and access the resources of a household for multilevel analysis [56]. where β is the weighted score of the basic amenities of the village and A is the score (presence absence or coded) of basic amenities of the village. A two-step algorithm was used to cluster the villages on the basis of the total geographical area of the village, population, the altitude of the village and remoteness index of the villages. Three clusters were formed on the basis of the 4 criteria given for analysis ( Figure 2). Cluster 1 comprises of 49 villages having high altitude, large geographic area, remote and less population. Cluster 2 comprises 11 villages having low altitude, easily accessible, moderate geographic area and high population. Cluster 3 again comprises of 49 villages having moderate elevation, small sized, moderately remote, sparsely populated. Climate change will increase the exposure of different villages to environmental hazards and will affect the agricultural sector more where irrigation facilities are limited. Conflict with wild animals and exposure to disasters will also increase due to climate shock. The villages having high dependency on agriculture, more prone to natural disasters and high wildlife conflict are more exposed to climate shock compare to the villages that lack those exposures. The village and household vulnerability also will be shaped by these factors. To include the factors associated with exposure to climatic stress in vulnerability analysis 102 villages from different clusters were selected for the primary data collection on the agricultural dependencies of the village, disaster proneness of the village and wildlife interaction. On the basis of this information another clustering was done using a two-step algorithm. A total of 10 sub-clusters were formed, 5 from each Cluster 1 and 3, (Table S2 of supplementary). Due to the smaller number of villages in Cluster 2 (11 villages) it remained as a single cluster. Thirty villages were selected representing all the clusters for detail household-level survey using household economy approach ( Table 1). The household economy approach (HEA) is developed by Save the Children, UK and designed to get information based on livelihood and access the resources of a household for multilevel analysis [56].   The number of households to be surveyed within a village is selected using standard queries. If the villages have less than 50 households, 50 percent of the households were surveyed, as the number increases to 100, 250, 500 and more than 500, 30 percent, 20 percent, 10 percent, and 5 percent of the households were surveyed, respectively. Within a village, households were selected through a stratified random selection process, accounting for the differences in the social and economic conditions of the households, gender and age structure of the respondents. household-level questionnaire surveys were conducted in 646 households of 30 villages (15 each from Uttarkashi and Tehri Garhwal district, India) using semi structured questionnaire. Information on the household economy, dependency level to agriculture, forest and other natural resources for daily needs and perceptions on climate hazards were recorded from each and every household. Information on different preidentified indicators following the deductive method [49] were collected. Indicators were classified under the four major capital viz. physical, human, financial and natural resources as part of the adaptive capacity of the households as most of the indicators of social capital like access to basic amenities already been estimated while calculating the remoteness index. The list of subcomponents and selected indicators under each capital are given in Table 2. Some of the variables where the variance was very high and which were having non numeric responses, ranked information was used. In case of the education and occupation of the respondents, the information was non numeric, so responses were ranked as per the lowest to highest education and occupation classes. Similarly, in the case of agricultural area and agricultural production, the variance in the responses was very high so the data were ranked to different production classes from low to high. Individual scores of the indicators were transformed using minimum-maximum rescaling transformation using the equation used in Hahn et al. (2009) [42] (Equation (2)). Transformed scores were added to get the aggregate score for each subcomponent of four major capitals. The aggregate score of the components depicted the innate adaptive capacity or resilience of the households for climate change vulnerability. The average resilience of each village was calculated, and village-level indicator values were added to calculate the village-level innate resilience score. Cluster analysis was done using R software including the values of different capitals along with the resilience scores, as the variability of the resilience score within the villages was very small (ranges from 8.105 to 13.386). Villages and households were classified into different classes from highly resilient to highly vulnerable. The ratio of between and the total sum of squares was compared for deciding how many clusters can better classify the villages and households in different classes. The agricultural dependency of the villagers, instances of wildlife conflict, exposure to disaster in past and its effect on the villages were taken as indicators to assess the level of intensity of the climate change shock on the villages. Agricultural dependency of the villages was categorized as high, medium and low/nil and scored as 3, 2 and 1, respectively. Exposure to natural disasters and wildlife conflict was also categorized and scored similarly. The sum of the scores was transformed using Equation (2) [42] and the normalized value as a risk score was deducted from the innate resilience score of the respective villages to get the vulnerability score of the villages.
Agriculture and forest are the two major areas to get affected by the effect of climate change in the near future. An estimated 4.9% reduction on the agriculture production for change in 1 • centigrade of temperature was projected by IPCC [57]. As natural and financial capital of the household and village largely dependent on the agriculture and forest resource of the villages, changes in the production and availability of these resources will definitely affect the vulnerability of the villages. Change in agricultural production in the future climate scenario was estimated for the villages to assess how changes in agricultural production will change the vulnerability score of the villages. For projections of future climate, Australian Community Climate and Earth System Simulator (ACCESS-01) jointly developed by the Bureau of Meteorology and Commonwealth Scientific and Industrial Research Organisation (CSIRO), with help from the Australian universities [58] was used. The "middle of the road" Representative concentration pathway 4.5 for GHG (Green House Gas) scenario is used. For analysis, the change in average annual temperature of current and the year 2050 time series climate change scenario (https://worldclim.org/cmip5_30s) is assessed. The future mean annual temperature (Bioclimatic variable 1) for all the villages was extracted using extract values to points of spatial analyst tool in ArcGIS 10.2. The present mean annual temperature and expected mean annual temperature of the villages were compared and the difference in mean annual temperature was multiplied by 4.9 (as 4.9% decrease of agricultural crop production is expected for per degree centigrade temperature increase) to get the percentage change in the future agricultural production. The percentage score for all the villages was transformed using Equation (2), and the transformed score was deducted from the vulnerability score to get the future vulnerability score of the villages due to loss of agricultural production because of the climate change effect. Excel, SPSS and R software are used for the analysis of the data.

Demography
Among the 646 respondents, 306 were females and 340 were male. As in the villages of the Indian Himalayan region household and agricultural work are driven mostly by women, it was tried that ratio of male and female respondents be the same. The difference between the number of male and female respondents was not significantly different statistically (p < 0.5 at 95% confidence level). The percentage of male and female respondents in each village surveyed is given in Table 1. Similarly, the age and occupation of the respondent are also important as the perception and experience of the respondent are important while getting household information. An attempt was made to get equal respondents from all the age classes. Respondents were grouped in six different age classes from less than 20 to more than 60 years of age and in between 4 classes were of equal age width of 10 years ( Figure S1 in supplementary material). There was no significant difference in the number of respondents in between the age classes except the less than 20 years age class (p < 0.5 at 95% confidence level) ( Table 3). Agriculture is the major occupation (35.34%) followed by Government service, private service and casual labor, 6.67%, 6.18% and 6.02%, respectively ( Figure 3).

Household-level Resilience
Resilience scores for household-level resilience ranged from 5.03 to 16.54 depicting high vulnerability to high resilience, respectively. Five different clusters were formed as the ratio of between and the total sum of square values were reaching the asymptote near the cluster size of 5 ( Figure S2 of supplementary material). Along with the resilience score, values of different capitals were also used for clustering as range of resilience scores do not varied much between the

Household-level Resilience
Resilience scores for household-level resilience ranged from 5.03 to 16.54 depicting high vulnerability to high resilience, respectively. Five different clusters were formed as the ratio of between and the total sum of square values were reaching the asymptote near the cluster size of 5 ( Figure S2 of supplementary material). Along with the resilience score, values of different capitals were also used for clustering as range of resilience scores do not varied much between the households. Mean values of resilience scores and different capitals for each cluster were given in Figure 4.  Household resilience status depicts the highest number of households were of resilient classes (31.57%) followed by vulnerable (22.29%), moderate vulnerable (18.575%) and highly vulnerable classes (17.49%). Only 10% of households fall in highly resilient classes (details are in Table 1

Village-level Vulnerability
Average household-level resilience score was calculated for each village and added to the normalized value of the village-level indicators viz: altitude, remoteness index, total geographic area, population, to get the village-level resilience score. Village resilience score along with the average of natural, physical, human and financial capital was used for clustering the villages in different vulnerability levels. Five different clusters were formed, and the range of each variable used for clustering in each cluster was given in Table 4 (supplementary materials Figure S3). Village-level resilience score is higher in Uttarkashi District (11.298 ± 1.35) than Tehri Garhwal District (9.965 ± 0.824) and the difference is statistically significant. The villages as per their vulnerable class are Household resilience status depicts the highest number of households were of resilient classes (31.57%) followed by vulnerable (22.29%), moderate vulnerable (18.575%) and highly vulnerable classes (17.49%). Only 10% of households fall in highly resilient classes (details are in Table S1

Village-level Vulnerability
Average household-level resilience score was calculated for each village and added to the normalized value of the village-level indicators viz: altitude, remoteness index, total geographic area, population, to get the village-level resilience score. Village resilience score along with the average of natural, physical, human and financial capital was used for clustering the villages in different vulnerability levels. Five different clusters were formed, and the range of each variable used for clustering in each cluster was given in Table 4 (supplementary materials Figure S3). Village-level resilience score is higher in Uttarkashi District (11.298 ± 1.35) than Tehri Garhwal District (9.965 ± 0.824) and the difference is statistically significant. The villages as per their vulnerable class are plotted in the altitudinal gradient of the Bhagirathi Basin in Figure 5a.   Among the five clusters, Cluster 1 represents seven villages (Baijkot, Bayana, Hitanu, Chamiyala, Kumrada, Indergaon, Srikot) with moderate resilience having higher financial and natural capital. Cluster 2 comprises of four villages (Pata, Barsu, Hadiyari, Purali,) with high resilience having higher natural physical and financial capital. Cluster 3 comprises of seven villages (Bhatwary, Garat, Pakh, Sitakot, Dharwal, Pabela, Musangaon) having moderate vulnerability because of high human capital, but lower scores for other capitals. Cluster 4 comprises of eight vulnerable villages (Dharali, Gajoli, Barsali, Koti, Kyal baghi, Pata (T), Ramgarh, Pipola) due to their very low physical and human capital and cluster 5 comprises of four highly vulnerable villages (Ladari, Khand, Girgaon, Gujetha,) due to their low financial and natural capital.
The vulnerability score deduced was then compared with the innate resilience score to group villages in different vulnerable categories on the basis of cluster averages of resilience score and the cumulative score of human, financial, physical and natural capital. Vulnerability status of 11 villages had changed after deducting the risk score ( Figure 6). Barsu and Hadiyari village which were highly resilient in innate resilience shifted to resilient and moderate vulnerable group, respectively after deducting the climate risk score. Bayana, Hitanu and Kumrada shifted from resilience to moderate vulnerable group. Dharali, Kyal Bagi and Pata village of Uttarkashi shifted from vulnerable to highly Among the five clusters, Cluster 1 represents seven villages (Baijkot, Bayana, Hitanu, Chamiyala, Kumrada, Indergaon, Srikot) with moderate resilience having higher financial and natural capital. Cluster 2 comprises of four villages (Pata, Barsu, Hadiyari, Purali,) with high resilience having higher natural physical and financial capital. Cluster 3 comprises of seven villages (Bhatwary, Garat, Pakh, Sitakot, Dharwal, Pabela, Musangaon) having moderate vulnerability because of high human capital, but lower scores for other capitals. Cluster 4 comprises of eight vulnerable villages (Dharali, Gajoli, Barsali, Koti, Kyal baghi, Pata (T), Ramgarh, Pipola) due to their very low physical and human capital and cluster 5 comprises of four highly vulnerable villages (Ladari, Khand, Girgaon, Gujetha,) due to their low financial and natural capital. The vulnerability score deduced was then compared with the innate resilience score to group villages in different vulnerable categories on the basis of cluster averages of resilience score and the cumulative score of human, financial, physical and natural capital. Vulnerability status of 11 villages had changed after deducting the risk score ( Figure 6). Barsu and Hadiyari village which were highly resilient in innate resilience shifted to resilient and moderate vulnerable group, respectively after deducting the climate risk score. Bayana, Hitanu and Kumrada shifted from resilience to moderate vulnerable group. Dharali, Kyal Bagi and Pata village of Uttarkashi shifted from vulnerable to highly vulnerable group. Same way Pakh, Pabela and Musangaon villages shifted from moderate vulnerable-to-vulnerable group (Figure 5b). Future agricultural production loss will also affect the vulnerability of the villages and the result shows 9 of the 30 villages shifted from present vulnerability class to higher vulnerability classes. Gajoli, Pakh, Barsali, Koti and Pipola were shifted from vulnerable to highly vulnerable class ( Figure  5c). Dharwal and Kumrada shifted from moderate vulnerable-to-vulnerable class and Purali and Chamiyala village shifted from highly resilient and resilient to moderate class. The villages under different resilience and vulnerability class in the different scenario were given in Table 5. Future agricultural production loss will also affect the vulnerability of the villages and the result shows 9 of the 30 villages shifted from present vulnerability class to higher vulnerability classes. Gajoli, Pakh, Barsali, Koti and Pipola were shifted from vulnerable to highly vulnerable class (Figure 5c). Dharwal and Kumrada shifted from moderate vulnerable-to-vulnerable class and Purali and Chamiyala village shifted from highly resilient and resilient to moderate class. The villages under different resilience and vulnerability class in the different scenario were given in Table 5.

Discussion
South Asia is home of about 600 million poor people of a total of about 1.5 billion people residing in and the number exceeds half of the world's total poor and marginalized people and their dependency on climate sensitive sectors like agriculture, forestry and other natural resources are high for the fulfillment of day-to-day needs [59]. As an extremely vulnerable region to climatic hazards 750 million people of South Asia were affected between 1990 and 2006 and experiences 230,000 deaths and damages that cost about $45 billion [60]. Countries in the Hindukush Himalayan region, includes India, Bangladesh, Nepal, Bhutan and Pakistan are facing increased frequency and magnitude of extreme weather events and more extreme weather events are likely due to climate change in future, and this will worsen the situation of South Asia over the next decades [61,62]. Extreme temperature reduces yields of agricultural crops and exposes the land for weed and pest proliferation whereas changes in the precipitation pattern increases the chance of crop failure and ultimately causes production decline in the long run and will ultimately threatens food security [63]. The overall impact of climate change on the agricultural sector is likely to be negative, although for some crops there will be gain in production to some specific regions to some extent [57]. As the global climate change shows its effect over the last two decades, sectors like biodiversity, human health, water and energy, agriculture and food security will get intense stress and will cause immense poverty and vulnerability in South Asia [59,62].
The Himalayan ecosystem is the lifeline of the people of South Asia, mostly the people living in the flood plains of the major rivers and their tributaries. The impact of climate change is predicted to change the flow of river water mostly during the dry summer season and expected to have large scale impact on irrigation, hydropower and other ecosystem services [62,64]. In the present study, the innate resilience of the households and villages of Indian Western Himalaya was assessed on the basis of the assets and capitals of the particular households. The difference in the resilience score and differences in assets to form different clusters are mainly based on natural, financial, physical and human capital. Although vulnerability or resilience of the villages depends on multiple components, identified as indicators and all the indicators act differently and additively for portraying vulnerability. All the indicators taken in the natural capital and most of the indicators of financial capital were directly dependent on nature and thus under the effect of climate related responses of different environmental parameters. On an average 64.66% of the financial capital value of the households are dependent on the climate variables. Production of all agricultural products are likely to be affected with change in temperature in future climate scenario. Hence, change in the availability of these parameters had changed the indicator values and had shifted the household and villages to the next vulnerable or resilient classes (Table 5). The villages which were highly dependent on agriculture and natural resources were prone to climate change impact and the vulnerability scores of the villages were also found low and grouped in vulnerable and highly vulnerable classes. It is evident from the result, that the villages of Tehri Garhwal are much more vulnerable. Human capital is found to be less in households and villages surveyed in Tehri Garhwal district. Due to the inception of Tehri Dam and the new township, most of the villages in the periphery of the dam lost the bulk of their agricultural area. Although compensatory schemes were given, but the one-time financial assistance for that inevitable loss of the land to the agriculture dependent community was not sufficient for generating enough livelihood options for the residents. The youth of those villages were compelled to leave for the nearest township for better livelihood resulting in the decrease of human capital for the households and the villages. Villages near the town and on the road connectivity to the several pilgrimage areas get some financial assistance from small scale businesses, but others deprived of the facility are under the shock of the developmental prejudice. The shock generated from the climate related hazards cannot be bearable by the households and the villages having less capital or assets and minimum livelihood options. Thus, the status of resilient household and villages changes as per their coping potentials. In the Indian Himalayan region, under the fragile landscape and the arduous geophysical condition livelihood option of the inhabitants are very much restricted and mostly dependent on the natural resources available to them. The change in the availability of natural resources and bioclimatic factors are thus crucial for their future livelihood strategies. The chances of getting other sources of income being there is negligible so far. The trend of outmigration for a better livelihood facility was also documented.
Strategies envisioned for future climate change adaptation should be developed with the full understanding of the complexities involved that causes the vulnerability at the present time [6]. The place and time specific configuration of three analytical components, the human ecology, expanded entitlement and political economy define the dimensions of the vulnerability. The dimension of vulnerability that determines the risk exposure, coping capacity and recovery potential is defined by the three components [6,19]. Vulnerability analysis helps to identify the vulnerable and resilient group of households or villages. It also allows the clear identification of whether there are other important factors that threaten the livelihood and resilience of the inhabitants except climate change. The analysis can assist in the designing of the safety nets or interventions that will help in the method of reducing the risk or improve the risk management capacity (resilience). As Vulnerability analysis is done using multiple indicators it is able to accommodate the multiple dimensions in terms of asset base, the flow of income, farming productivity, access to social and government support, resources and services at the households, etc. Thus, the profiling of vulnerable groups can be built and factors can be identified for improving the reliability of safety nets [7]. The present study gives a replicable analytical framework to use different indicators for analysis of innate resilience and vulnerability of villages in the Indian Himalayan Region. The process of evaluation of vulnerability and resilience can be followed for assessing the same at a landscape level and will be helpful for future administrative and policy level interventions. Hence, the result of the present study has manifold implications on the policy level to decide where the assistance is needed, how much is needed, and for how long as per the framework of household economy approach.

Conclusions
Before vulnerability can be addressed, it is vital to identify who is vulnerable and to what extent. Vulnerability assessment allows investigation of the complex relationship of socioeconomic and demographic factors that are being impacted by different climate stressors. The Indian Himalayan region is socioeconomically vulnerable to climate change. Imbalance in wealth distribution, remote rural population and dependency on agriculture for the economic stabilization enhance the chances of vulnerability. The index-based approach used in the present study is useful to assess both the impacts and the societal capabilities to adapt to climate change effects. The present study has explored the analytical utility of using index-based assessment of adaptive capacity and thus provides an assessment tool that can be used on a local scale and assess household and village-level adaptive capacity. The methodology can also be adopted to compare and contrast adaptive capacity and vulnerability in larger spatial scale and different socioeconomic settings. The study identifies both the villages and the factors responsible for the resilience and vulnerability that can be used and integrated into microplanning and macro-policy development for better resource allocation to the vulnerable community as part of the climate change mitigation and adaptation planning.
Supplementary Materials: The following are available online at http://www.mdpi.com/2071-1050/12/17/6938/s1, Figure S1: Age-class distribution in the respondents, Figure S2: Ratio of between and Total sum of squares of cluster numbers, Figure S3: Range of each variables used for village clustering, in each cluster (a) village resilience score; (b) Human capital; (c) Natural Capital; (d) Financial capital; (e) Physical capital, Table S1: Households in different clusters, Table S2: Villages of Uttarkashi and Tehri Garhwal districts under different cluster and sub-clusters.
Author Contributions: S.D.: Conceptualization, methodology, investigation, formal analysis, writing-original draft preparation; R.B.: conceptualization, supervision, methodology, visualization, writing-reviewing and editing, project administration. All authors have read and agreed to the published version of the manuscript.