Version of Record: https://www.sciencedirect.

com/science/article/pii/S2212420919308878
Manuscript_c56a27e5fe1a89cbd46fe45b573ef735

2 An Institutional Analysis of Glacial Floods and

3 Disaster Risk Management in the Nepal Himalaya
4 Ian Thompson1, Milan Shrestha*1, Netra Chhetri2, Datu Buyung Agusdinata1
5
1
6 School of Sustainability
7 800 S Cady Mall
8 Arizona State University
9 Tempe, AZ 85281-5502
10 USA
11

12 2School for the Future of Innovation in Society

13 1120 South Cady Mall
14 Tempe, AZ 85287-5603
15 USA
16

17

18 *Corresponding author: milan.shrestha@asu.edu

19 Tel: +1-480-965-7175
20 ORCID: 0000-0003-4758-515X
21

22

23

24

25

26

27

28

29

30

© 2020 published by Elsevier. This manuscript is made available under the Elsevier user license
https://www.elsevier.com/open-access/userlicense/1.0/
31

32 Abstract
33 Institutional factors of disaster risks have rarely been studied for glacial lake outburst floods
34 (GLOFs), which are threatening high mountain regions such as the Andes, Alps, and Himalayas.
35 Exacerbated by climate change, much of the focus in the Himalayas so far has been on GLOF
36 risk anticipation. Recent research suggested the gaps in how scientific information on GLOFs
37 has been communicated to local communities and the resulting distrust in government-led
38 mitigation efforts such as, lake lowering and early warning systems. These gaps are largely
39 driven by a set of socio-economic, cultural and institutional factors, which are poorly
40 understood. Applying an institutional analysis framework, this research examines the
41 institutional conditions under which Himalayan communities can create effective strategies to
42 address GLOF risks. This research applies a mixed-methods approach by combining a)
43 quantitative analysis of household social surveys, and b) qualitative analysis that maps out the
44 various institutional factors that influence the success of community-level mitigation efforts.
45 Additionally, GLOF case studies from Nepal are compared to those of Peru, where a long history
46 of managing GLOFs exists. A set of critical institutional factors and barriers have been identified,
47 including lack of cross-scalar communication networks, a mismatch between the agenda of
48 external agencies and priorities of local communities, failure to recognize the value of local
49 knowledge and participation in policy processes, and ineffective inter-organizational
50 coordination of knowledge sharing and funding streams for mitigation projects. Bridging these
51 disconnects is a significant step toward successful adaptation and mitigation efforts to reduce
52 future GLOF disaster risks.
53 Keywords
54 Institutional Analysis, Disasters, Sustainability, GLOF, Himalaya
55

56

2
57 1. Introduction
58 Glacial lake outburst floods (GLOFs) are threatening high mountain ranges such as the Andes,
59 Alps, and Himalayas. GLOFs and similar glacial flood hazards are responsible for thousands of
60 human casualties and severe destruction of critical infrastructures worldwide (7, 46, 10, 24).
61 Although mountain communities have been dealing with GLOF risks for centuries, climate
62 change is exacerbating both physical and social vulnerability (20, 24). Science behind the causes
63 and characteristics of these glacial hazards is well-known, but less well-known is the socio-
64 economic, cultural and institutional drivers behind creating effective strategies to mitigate
65 GLOF risks (8, 9). The institutional factors driving disaster risks have rarely been studied in the
66 Himalayan region, as much of the focus so far has been on improving GLOF risk anticipation and
67 other scientific inquiries.
68 GLOFs can be catastrophic, as increasing water volume in glacial lakes from glacial melt puts
69 pressure on unstable moraines and can be further triggered by avalanches, rock/landslides, and
70 earthquakes leading to increased frequency of flood events. Extreme weather events such as
71 heavy rainfall, blizzards and landslides may accelerate the occurrence of GLOF events (51, 38),
72 posing a serious risk to the safety of downstream communities, infrastructures, and their
73 livelihoods.
74 Our goal in this paper is to analyze the institutional conditions under which Himalayan
75 communities interact to create effective adaptation and mitigation strategies for climate-
76 induced hazards. We are guided by the following research questions: a) How do the Himalayan
77 communities respond to the GLOF hazards and risks? b) What rules and institutions are in use
78 as part of their response to deal with such hazards and risks? and c) How effective have those
79 institutions been, given the context that more formal institutions (e.g., government-controlled
80 projects, mitigation options proposed by outsiders) are now replacing the traditional
81 institutions (e.g., customary rules of resource management, faith-based relief initiatives)?
82 An institutional analysis, broadly defined as a systematic assessment of both the rules and
83 structures that govern interactions of people (14, 30, 32), is a widely used framework to analyze
84 how local communities perceive, interpret, and react to government policies. In this research,
85 an institutional analysis framework is applied to directly respond to the key findings of Sherpa
86 et al (40), which highlighted the gaps in how scientific information on GLOF and other glacial
87 hazards has been communicated to local communities in the Sagarmatha (Mt. Everest) Region
88 of Nepal (40). By applying this framework, we seek to disentangle the reasons behind the local
89 residents’ distrust of government-led mitigation projects such as, lake lowering and early
90 warning systems. Bridging these gaps is a significant step toward successful mitigation efforts,
91 but it is also important to establish that they are essentially driven by a set of socio-economic
92 and institutional factors, which are poorly understood.
93 Although institutional analysis of GLOF disaster risks is rare in the Himalayas, there is much to
94 be learned from the Peruvians, who have actively been dealing with GLOF disasters as early as

3
 95 the 1940s. Over time they have developed sophisticated geo-glacial engineering expertise,
96 infrastructures, and institutions (8, 5). A case study by Carey et al. (8) analyzes adaptation
97 efforts for Lake 513 in the Cordillera Blanca range in Peru by using a socio-environmental
98 framework. The Peruvian Andes case provided an in-depth look at the Social-Ecological Systems
99 (SES) factors that either facilitated (i.e., technical capacity, visible hazards, institutional support,
100 committed individuals, international role) or impeded (i.e., perceived risks, imposed policies,
101 institutional instability, knowledge disparity, invisible hazards) attempted response strategies
102 for this specific glacial lake (8). They offer a valuable holistic approach for disaster risk reduction
103 in high mountain ranges versus traditionally engineering-based analyses.
104 Our focus on the institutional analysis of a representative case of disaster risk management and
105 governance is related to the Imja Tsho (lake), a supra-glacial lake categorized as one of the most
106 dangerous lakes with GLOF risks in Nepal (19, 11). Previous studies of Imja Lake reported that
107 the lake poses a high GLOF risk to downstream communities (19, 50 for historiography). Given
108 the growing media attention, the Government of Nepal (GoN) started to prioritize mitigation
109 efforts in 2014 to reduce GLOF risks from Imja Lake, including the integration of Early Warning
110 Systems (EWS) into several downstream communities (45, 11). The Watanabe et al. (50) study,
111 in particular, provides a useful look into the complexity surrounding the Imja Lake GLOF risks,
112 mainly cautioning that some studies overplayed the GLOF risks and by doing so, would be
113 taking attention away from potentially more dangerous hazards in the Nepal Himalaya (50).
114 However, other studies since Watanabe et al. (2009) have shown that Imja Tsho still poses risks
115 to local and downstream communities. Several others argue that while a catastrophic flood
116 event from Imja Tsho lake is unlikely, the high volume of water and rapid expansion of the lake
117 remained hazardous, which called for continued monitoring of the Imja Tsho (37, 42, 45). Such
118 controversial findings run the risk of confusing affected communities, especially as Imja Tsho
119 was ranked as an immediate risk under Nepal’s National Adaptation Plan of Action (NAPA).
120 Conflicting scientific findings over the past few decades on potential GLOF risks also makes it an
121 interesting case to look at it from institutional perspectives. As shown by Terpstra (44), the
122 presentation of incorrect scientific data to communities can cause long-term distrust of local
123 governments and authorities, which in turn, can have lasting effects on the success of disaster
124 risk management projects. This phenomenon is further validated through cases such as Tsho
125 Rolpa of the neighboring Rolwaling Valley (13, 41) and various historical cases in the Peruvian
126 Andes (7).
127 Given the importance of good governance in disaster risk, this research applies an institutional
128 analysis framework of disaster risk management projects and their interactions with local
129 communities to examine the institutional conditions under which mountain communities
130 create effective adaptation strategies to address climate-induced hazards. More specifically,
131 the research seeks to understand the key institutional factors that influence disaster risk
132 reduction strategies within the Himalayan region. The following sections explore the region
133 under study where fieldwork was conducted by the research team in 2016 and 2017. Key
134 theoretical foundations in resilience and adaptation are then considered, and gaps in the
4
135 current literature are highlighted as key points to address in the discussion surrounding the
136 disconnect between institutions at different scales. The indicators discussed below frame the
137 quantitative analysis of social factors, as well as the gaps within Himalayan disaster risk
138 institutions.
139 2. Study site
140 This research focuses on the Sagarmatha (Mt. Everest) National Park and Buffer Zone (SNPBZ),
141 widely known to outsiders as the Mt. Everest or the Khumbu region. Currently located in the
142 Khumbu-Pasang Lhamhu Rural Municipality, this covers approximately 1148 km2 and is defined
143 by its rugged landscape where elevation can increase dramatically over a short distance (15).
144 High mountain environments of this region have shaped both the cryospheric ecosystem and
145 the livelihoods of Sherpa people, the predominant ethnic group within the park (43). Within
146 SNPBZ there are 63 settlements and a population of approximately 8,000 people. Livelihoods
147 are historically based in agro-pastoralism, taking advantage of the mountain environment (16,
148 43). Since the creation of the Sagarmatha National Park in the 1950s, livelihoods have been
149 transitioning to tourism which has now become a dominant source of income. Agro-pastoralism
150 also remains an important source of livelihoods, however. This is especially so for marginalized
151 groups who are in the periphery of the tourism industry (16).
152
153 Imja Lake is located at 5010m altitude in the north-eastern corner of the study site (Figure 1),
154 and its feeder glaciers are Lhotse, Lhotse Sar, and Imja. Only a small pond in the 1960s, Imja has
155 since expanded in both volume and size. With over 75 million cubic meters of water, Imja is
156 seen as one of the largest glacial lakes in Nepal (36). This rapid expansion, amidst conflicting
157 scientific assessments of outburst risk for the lake, has led to confusion in at-risk communities
158 and a useful example for an institutional response.
159
160 < Insert Figure 1 about here>
161
162 Imja lake has been under observation by remote sensing as well as field site surveys due to
163 concerns over its rapid expansion (45). As part of Nepal’s NAPA, communities were engaged in
164 Local Adaptation Plans for Action (LAPA) planning processes in 2013 to provide priorities for
165 reducing risk. A key recommendation of this was to lower the lake at least by 3.5 meters to
166 reduce outburst risk, which is similar to the emergency remediation work done in Tsho Rolpa in
167 2000-2005. While the design for the drainage system was completed in early 2015,
168 administrative procedures and failed bidding processes led to delays in construction until June
169 2016 and it was eventually completed in October 2016. With the project cycle ending in 2017, it
170 is unclear whether other elements of the project, particularly the capacity building initiatives at
171 community levels were effectively implemented as identified by the original planning process
172 (12).
173
174

5
175 3. Theoretical foundations and gaps
176 While Himalayan communities became aware of potential GLOF hazards and risks in recent
177 decades, it is important to recognize that mountain communities have always faced challenges
178 to make a living and have adapted to those challenges. Verticality, a concept championed by
179 John Murra (1968) and later applied in the Himalayan region by Rhoades (2007), describes how
180 the diverse ecosystems created by the topographic extremes have shaped the cultural
181 development of the communities that occupy the land (27, 34). Himalayan communities have
182 historically developed place-based and niche-based agricultural and pastoral practices to take
183 advantage of the ecological diversity within the region. Their adaptation measures included
184 livelihood diversification, which essentially complemented farming and livestock with the
185 income from regional trade. Because of this mountain-specific adaptive practice, the land itself
186 plays an integral role in the identity of these communities, creating a strong sense of place for
187 specific regions. Additionally, because of the regional isolation of mountain communities,
188 strong local institutions have formed to support their mountain specific way of life, which in
189 turn, has fostered adaptation strategies to deal with natural hazards and risks that are endemic
190 to the region (43). The social systems that define the Himalayan communities are shaped by the
191 cryosphere and related ecological systems in which they are located. It is in this sense,
192 livelihoods such as agropastoralism and tourism are also closely linked to glacial lakes that may
193 also be the source of risks.
194 To understand the current gaps in disaster risk preparedness institutions in Nepal, we build on
195 concepts of community-based disaster risk reduction (39) and community-based resilience (3).
196 Concepts such as agency, self-organization, and local participation in policy provide a lens to
197 analyze institutional efforts and understand how social factors such as livelihoods influences
198 risk perception in the SNPBZ region by building on the work of Sherpa et al. (40). These
199 concepts parallel those in sustainability such as equity and inclusion, an approach that
200 recognizes complex adaptive systems, especially with regards to understanding how disaster
201 response strategies are shaped (22). Sustainability is a topic of growing attention in
202 international communities, especially after the adoption of 17 Sustainable Development Goals
203 (SDGs) in September 2015 by the United Nations. These goals play an active role in influencing
204 policies and programs surrounding climate change adaptation and hazard risk management in
205 Nepal (28). Additionally, the team draws upon institutional development in Peru which has
206 faced similar barriers to developing effective adaptation efforts that are inherent to complex
207 adaptive systems and the complexity of mountain development (8, 9).
208 This paper builds on the existing literature on glacial floods and other mountain hazards by
209 positing that current institutional efforts are doing little to incorporate these lessons into
210 adaptation strategies. The case for this will be built in the following sections, comparing several
211 institutional cases in which the neglect of various social elements led to failed efforts to
212 minimize risk.
213

6
214 3.1 Social Perceptions
215 While the general risk of GLOFs is recognized by Himalayan communities, the degree of risk is
216 influenced by several socio-economic factors. Understanding these factors is vital in
217 understanding the various ways the communities may react to the risk emanating from GLOF
218 events and as well as take measures to adapt to them. Dahal & Hagelman (13) reveals how
219 different factors such as previous exposure to GLOF events, “cry wolf” effect from failed Early
220 Warning Systems (EWS), and types of livelihoods shape the perception of risks held by
221 community members regarding GLOF hazards. The authors provide a qualitative analysis that
222 highlighted several social factors influencing how community members perceive the risk of
223 GLOF events that are reported by government officials and research institutions. Factors such
224 as cognitive biases and heuristics were shown to have substantial impacts on community
225 perception of risk to GLOF events that often contradicted the scientific reporting on the risk of
226 GLOF in this case of Tsho Rolpa Lake (13).
227 After several evacuations, and the absence of GLOF occurrence, the community began to
228 distrust both the mitigation measures put in place and the institutions behind risk monitoring.
229 An example is the evacuation of several at-risk communities in the Tsho Rolpa floodplain in
230 1997 without the occurrence of a GLOF. Following this event, researchers reported an
231 increasing distrust in government officials, scientists, and NGOs. Additionally, Peruvian
232 engineers in 1988 and 1989 rejected implementing early warning systems for Lake 513 with the
233 possibility that it would further increase community distrust in future government projects (8).
234 This shows the criticality of social perception in adaptation strategies had a lower perceived risk
235 as a result of public distrust.
236 3.2 Institutions and governance of GLOF disasters
237 In the context of this paper, institutions refer to the rules and norms surrounding disaster risk
238 management efforts (14, 30, 32). Defined as such, institutions expand beyond individual
239 organizations and focus on the processes of how decisions are made as influenced by various
240 social and cultural factors both formal and informal. For instance, formal institutions may
241 operate in terms of regulation and policy processes, while informal institutions operate within
242 cultural and spiritual norms and historical practices that have shaped adaptation strategies.
243 Regarding GLOF management, Peru has the longest history of institutional development that
244 focused specifically on reducing risk of GLOF events (7). Disaster risk reduction was triggered by
245 the 1941 GLOF of Laguna Palcacocha that led to 5,000 deaths and the destruction of nearly
246 one-third of the provincial capital, Huaraz. In response to this tragedy, the Government of Peru
247 initiated the study of glacial hazards through scientific inquiry and monitoring to better
248 understand the threat at hand. Over the next few decades, the approach to GLOF management
249 continued to shift and transform which led to several adaptation strategies implemented with
250 varying degrees of success.

7
251 Failed implementation strategies were often the result of Peru’s hazard risk management failing
252 to recognize the importance of various socio-economic factors. According to Carey (7), one of
253 the barriers was the process of data collection and dissemination of the results. For several
254 decades, glacial studies were performed by international scientists who reported the results to
255 the national government in the Capitol, Lima. Such methods excluded the local at-risk residents,
256 which led to increasing fear and uncertainty regarding the lakes looming above them. It also
257 added to the historical regional disparity between the capital and the mountain communities
258 that were seen as “backwaters” (7). Additionally, many scientific studies conducted in
259 partnership with external interest groups in succeeding decades on potential GLOF hazards
260 were centered around generating hydroelectricity in the region. Andean communities
261 perceived that the national interest was more toward developing large scale infrastructure
262 projects to support urban centers rather than the safety of its mountain communities. This
263 mirrors the natural exploitation of mountain regions seen in the Himalaya regions that will be
264 discussed more in-depth below (34).
265 The historical development of GLOF management in Peru reveals several key insights. First, the
266 national government created programs that were designed to reduce the vulnerability of
267 mountain communities to GLOF management. While the national initiative helped generate the
268 science of glacial hazards, the absence of local engagement in the process deepened historical
269 divides between communities and the national government. This is largely due to a lack of
270 recognition of various socio-economic drivers, such as economic class and regional power
271 dynamics that shaped GLOF risk perception. Second, lack of transparency, added with the lack
272 of involvement of community leaders in the planning process led to social resistance to
273 attempted projects in several at-risk areas. Local buy-in was not established and there was no
274 attempt to incorporate indigenous knowledge into project designs. This could have streamlined
275 prioritization of mitigation projects based on community input, while simultaneously building
276 trust with communities by incorporating their concerns. Third, agenda-setting played a public
277 role in the prioritization of several glacial development projects. An example was under
278 President Fujimori who promoted a neo-liberalization of many public works projects in Peru.
279 During this time, most glacial management projects were turned over to ElectroPerú, who was
280 a partner with American-based Duke Energy (7). Aimed at increasing water volume for
281 hydroelectric production of several lakes, the government of Peru also initiated several
282 contested projects. Communities had petitioned to have several of these lakes drained, and
283 public resistance to such projects caused them to be shut down. While there is much to learn
284 from Peru’s handling of GLOFs, Nepal remains distinct in various dimensions. For instance,
285 concepts of Buddhist Dualism are integral in explaining events as both natural and supernatural
286 in SNPBZ communities (2). This thought process strongly mediates how mountain communities
287 with strong a Buddhist culture view risk from natural events. Additionally, the structure of local
288 decision-making processes regarding natural resource management differs between Peru and
289 Nepal. Nepalese mountain communities have a long history of collective governance through

8
290 traditional systems based on the ideals of community-based action, and now more formal local
291 institutions such as community user groups (49).
292 3.3 External Institutions
293 External institutions consist of government and other non-governmental institutions that
294 articulate the community’s voice concerning power relationships, responsibility and
295 accountability. They operate at multiple scales with each operation is based on organizational
296 objectives and outputs that guide decisions in GLOF related projects. In the case of Nepal, these
297 institutions include the GoN, UNDP, and several international cooperation, leading to the
298 evolution of a complex institutional landscape to reduce the risk of GLOF events. Furthermore,
299 the diversity of institutions creates the multitude of agendas and priorities and drives a
300 complex decision-making landscape. Such diversities create complications in ensuring effective
301 policy-to-action efforts, funding, and communication networks with local communities that are
302 prioritized for reducing the GLOF risks. While these challenges are clearly articulated in
303 adaptation policy documents such as LAPA, there is a lack of transparency among the
304 institutions (6), creating several challenges for effective participatory approaches. One such
305 challenge is for external institutions to provide continuity in cross-scale communication and
306 consistent funding mechanisms for proposed adaptation strategies. For instance, communities
307 often report infrequent reporting by research organizations, and many projects remain
308 unfunded or with no clear sources identified (35). Adding to this, many of the institutional
309 partnerships happening at national and international levels to create policy lack any sense of
310 ownership at local levels. While the knowledge generated from such efforts offer valuable
311 information for creating policy, the absent of user input makes it difficult to translate into
312 action. Shaw highlights the importance of including community input into policy efforts to
313 foster local adaptation capacity and address disconnects between decision-making at each
314 scale (39). This emphasizes the importance of continuous community-input for policy-making
315 processes to ensure successful community-based disaster risk reduction.
316 In addition to policymaking and research institutions, there are several national and
317 international Non-government Organizations and Civil Society Organizations involved in risk
318 reduction in Nepal (4). From 1997 to 2014, over forty thousand of these organizations were
319 registered in Nepal (21). These organizations certainly play a role in both disaster relief and
320 local development benefitting from a more decentralized project management approach,
321 compared to the central government. Because of this, they are often lauded for being more
322 representative of local interests. NGOs also have drawbacks as they are often from outside of
323 the local social and cultural fabrics. They often operate independently of other organizations,
324 creating wasteful overlap of certain functions, and gaps in information dissemination (4).
325 3.4 Local & Informal institutions
326 Mountain communities have a long history developing effective local institutions and creating
327 strategies fit for their circumstances (34). For example, across the Khumbu valley, to manage

9
328 common-pool resources such as forest and rangelands, a historical customary rights approach
329 known as “shingha nauwa” was prevalent (23, 43). Operated through local community efforts,
330 this self-organized system is credited for the collective governance of local forests. Such
331 customary systems that operated based on local experience and knowledge system considered
332 the nuances of the cultural composition of each involved community (1). Through this,
333 communities were able to establish rules-in-use that dictated seasonal harvesting of resources,
334 as well as social-enforcement mechanisms and leadership hierarchies (23). This reveals that
335 social cohesiveness through the trust was an essential driver of creating effective localized
336 management practices. With the establishment of the SNPBZ, customary rights to collect forest
337 resources were superseded by the rules of national park's conservation policies, effectively
338 displacing communities’ control over the resources managed at the local level for generations.
339 Two critical factors resulted in this regime shift. First, as by displacing customary resource
340 management systems, mountain communities were treated as the peripheral to the political
341 power. Second, the national government was able to impose, with the influence and support of
342 international organizations, more formal rules of resource management on that local
343 communities that were perceived by the residents as more exploitative (49).
344 With the realization that centralized management approaches were not effectively conserving
345 forest resources, the government transitioned back to a community-based management system
346 by implementing a community forestry program in the early 1990s. While this was done in the
347 realization of the importance that participation plays ineffective governance of natural
348 resources, it also ignored key social dynamics that have created contention with communities.
349 For instance, rather than taking advantage of traditional local institutions (e.g., shingha nauwa),
350 the national government created community forest user groups in each region under the
351 District Forest Offices (DFOs). While it is notable that this was done in consideration of equity in
352 communities, these efforts were largely influenced by external agents. This created
353 disconnected institutions by ignoring existing social cohesion and creating a new power
354 relationship through the management of new institutional leadership (49). Because of this,
355 effective institutional management across scales become convoluted, as the national agenda
356 pushes participatory planning considering competing interests at local levels (49).
357 Another critical element of analyzing institutions within the case of Nepal is how community-
358 based disaster risk reduction will take place. Shaw (39) highlights the need for the consideration
359 of local capacities and their roles in reducing localized risk in several cases throughout Asia.
360 Some of the more critical points mentioned include that strong local institutions are vital to
361 create effective resource governance regime at local level and should be integrated into
362 policies to ensure synergistic efforts. Similar institutions architecture evolving specifically for
363 the Mt. Everest region. Engel et al. (17) provide useful insight in how subcultures form in
364 regions that have historically experienced disasters. In such regions, natural disasters have
365 played key roles in shaping a culture around the necessity of adaptation and mitigation
366 strategies. The authors also emphasize that disaster management should happen within the
367 cultural frameworks of the subcultures that have formed within disaster-prone regions. While
10
368 this study is framed to subcultures formed in the Netherlands, such concepts prove especially
369 relevant to those in Nepal, in which verticality and the unique social-ecological system have
370 formed distinct subcultures across various Himalayan communities (34).
371 <Insert Figure 2 about here>
372 Figure 2 shows the institutional framework as within the Buffer Zone under the Department of
373 National Park and Wildlife Conservation (DNPWC). Here, User Groups (UG) manage resources
374 through community-based approaches with the Committee representing community leadership
375 in three districts (Namche, Khumjung, Chaurikharka). Another local institution is represented in
376 the Sagarmatha Pollution Control Commission (SPCC), which is a bottom-up management
377 system for waste collection and disposal within the SNPBZ. With the addition of both trekking
378 and lodging associations, there is a diverse representation of local interests in which
379 communities have self-organized to manage decisions. This links back to the concept of
380 verticality, in which the heterogenous SES conditions requires a diverse management strategy
381 on a regional scale. This is problematic when considering that adaptation strategies for
382 regionally and culturally specific challenges are originating from external institutions that are
383 driven by disconnected political and economic interests. This is especially relevant with concern
384 to livelihoods, in which adaptation strategies must recognize the importance of how they are
385 shaped by the environment as well as their impact on risk perception (40). While the changes in
386 the national political landscape, creation of the Sagarmatha National Park, and growth of the
387 tourism industry have transformed the institutional landscape in the region, local institutions
388 are still vital to decision-making processes within the region.
389 3.5 community-based adaptation

390 The discussion above highlights the need for context-specific solutions to addressing
391 community risk. One such framework is the community-based disaster risk reduction (CBDRR)
392 that outlines several conditions for effective CBDRR (39). Local institutions (formal and
393 informal) play a critical role in maintaining community initiatives. Shaw (39) also emphasizes
394 the need for local initiatives to be integrated into government policy to ensure scalability.
395 Additionally, local change agents play a critical role in implementation at local levels, backed by
396 the compatibility with development policy. The CBDRR process also outlines several factors
397 such as awareness by planning, vulnerability assessment, and risk management practices at
398 local levels that lead to more sustainable outcomes. An example of this type of effort briefly
399 described above is through the LAPA process. The LAPA is a participatory planning approach
400 that seeks to develop community adaptation plans to reduce risk to GLOF events and climate
401 change. Byers and colleagues (6) show how this approach is being integrated into both Peruvian
402 and Nepalese institutions in partnership with the High Mountains Adaptation Partnership
403 (HiMAP). Through providing a “science-based, community-driven approach” to building
404 adaptive capacity, LAPA seeks to address historical institutional asymmetries such as lack of
405 transparency that have led to community distrust to government adaptation strategies. While
406 these efforts vastly improve upon stakeholder engagement and integrating local knowledge,
407 several factors impede institutional capacity for implementing these projects at a larger scale.

11
408 Peniston (33) shows that shortcomings such as regional scalability, effective coordination for
409 infrastructure and funding streams and inter-organizational cooperation are all factors that will
410 need to be addressed for LAPA to have serious impacts on climate change adaptation and
411 hazard risk management processes.
412 This paper also draws upon the works of Berkes and Ross (3) who lay out several conditions for
413 building community resilience through the integration of ecology and psychology concepts.
414 Through this, community resilience is composed of agency and self-organization, with core
415 concepts emerge such as people-place connections, values and beliefs, knowledge, skills and
416 learning, social networks, engaged governance, community infrastructure, leadership, and
417 positive outlook. These factors, therefore, offer tools for evaluating the efficacy of projects that
418 are implemented by external institutions such as the Government of Nepal and aid agencies.
419 These also play a strong role in the institutional analysis provided in the results and the
420 discussion below.
421

422 4. Methods
423 This research employs mixed methods (quantitative and qualitative) approach to analyze the
424 current institutional landscape for disaster risk management in Nepal. The main dataset used in
425 this research is drawn from the household social survey carried out by a multidisciplinary
426 research team in 2016, which collected 138 household surveys in nine different villages of the
427 Khumbu and Pharak valleys of the SNPBZ using a stratified random sampling technique and a
428 survey instrument administered in Nepali language. A subset of the household social survey
429 data was analyzed to answer questions regarding coping mechanisms and institutional capacity
430 (see Appendix A). A key underlying philosophical assumption of the study is that the
431 subcultures formed in disaster-prone areas have agency to implement effective adaptation
432 strategies, and yet, lack the resources to coordinate at larger scales and are subjected to
433 agenda-setting by external institutions (see 17).
434 4.1 Quantitative analysis
435 To understand the links between risk perception, access to GLOF warning systems, and climate
436 change information, the study starts with an exploratory analysis using descriptive statistics. To
437 expand upon this, a logit model (Equation 1) was used with GLOF risk perception as the
438 independent variable (Y), and sources of climate change, or GLOF warning system factors as the
439 dependent variables+:
440 Y = β + β x + β x + …..
0 1 1 2 2

441 Additionally, Classification and Regression Tree (CART) analysis was conducted with GLOF risk
442 perception as the dependent variable, and the various institutional factors (based on questions
443 in Appendix A) as the independent variables. The analysis of the predictor variables reveals
444 which factors influence risk perception the most.

12
445 4.2 Qualitative analysis

446 Qualitative analysis consisted of document analysis of reports by various organizations involved
447 in GLOF risk reduction. Over 40 documents were reviewed, composed of: government reports
448 at each scale; disaster risk reduction case studies spanning the Himalayas, the Andes, and the
449 Alps; and theoretical frameworks to understand the socio-economic, cultural, and ecological
450 factors affecting institutional decision-making at each scale.

451 Key themes were identified as influencing adaptation strategy success and analyzed using
452 frameworks such as community-based disaster risk reduction (39) and community-based
453 resilience (3). Such themes presented in established frameworks are supported by key findings
454 in the various case studies and reports that emphasize the disconnect between institutions at
455 different scales, and the barriers they present.
456 The following questions were used to guide the analysis:

457 • Who are the major actor groups influencing disaster risk reduction in the Khumbu
458 region?
459 • What are their development and objectives and priorities?
460 • What are their funding streams, decision-making frameworks, and knowledge
461 dissemination pathways?
462 • What are the inter-organizational connections across scales?
463 The information collected was then used to create an institutional map for involved
464 organizations that incorporates decision-making at different scales, information channels, and
465 funding streams.
466

467 5. Results and discussions

468 The results of the household social survey provide some in-depth insights into how local
469 resident’s risk perception on disaster risk and mitigation measures is influenced by their
470 sources of climate change information, warning systems, and coping mechanisms. These results
471 are expanded upon by outlining gaps at different scales in the institutional landscape that
472 shape adaptation strategies at the community level.
473 5.1 Quantitative Assessment
474 When inquired about coping responses, 54% of respondents reported that they did nothing in
475 response to flooding events. This result indicates a lack of individual capacity to effectively
476 mitigate risk and loss given occurrences of floods. Such lack of capacity correlates with the
477 dissemination of climate change and flood warning systems in place. Approximately 31% of
478 respondents who have heard of climate change received information from static sources such
479 as radio, television, and newspaper outlets. It shows the predominance of one-way information

13
480 flows comparatively to the 11.6% who have learned through interaction with NGOs,
481 government agencies, and training experiences, and the 13.8% who have received information
482 through the education system. A similar trend can be seen from the distribution of flood
483 information sources. The largest source of flood information is shown to be reactive, through
484 sight and sound (30.4%) to know that a flooding event is occurring, compared to 16.7% of
485 respondents being informed through Early Warning System sirens, and 19.6% through mobile
486 device alerts. Figure 3 highlights the importance of warning systems, as those who identified no
487 warning system had the lowest (coefficient: 0.139) perception of risk.
488 <Insert Figure 3 about here>
489 Regarding the perception of responsibility for adaptation strategies, most respondents
490 indicated that the government should be responsible for both leading and funding projects. For
491 instance, 63% of respondents perceived that government officials should be leading projects
492 concerning GLOF hazards, and similarly, 66.7% indicated that the government should be
493 responsible for project funding, and 5.1% indicated that international agencies should play a
494 role in the process. While the perceived responsibility relies on the government, there is also
495 evidence of social restrictions for favored intervention strategies. While hazard zone planning is
496 often considered a favorable approach from a governmental standpoint (7), livelihoods and
497 geographic preference can highlight strong aversion towards potentially effective strategies.
498 For instance, when asked whether respondents would be willing to move their business to a
499 different location, 71.7% answered no. It shows both the importance that livelihoods play
500 regarding risk perception and the expectations put on government agencies to provide effective
501 strategies within the social structure of at-risk communities.
502 Further complicating the nexus of government onus and livelihood influence is in the parallel
503 between increased risk perception and livelihood generation. The top three natural threats
504 stated by respondents were water, scarcity, flooding, and blizzards. The major barriers to
505 tourism are stated to be physical: infrastructure, transportation, and difficult terrain (38%), and
506 other reasons such as bad weather, earthquakes, and lack of awareness. Many of these barriers
507 also overlap with disaster risk response which is especially challenging in remote mountain
508 regions as described above. This reiterates that there are multiple SES challenges presented to
509 mountain communities. Additionally, external institutions are faced with the burden of
510 providing effective management strategies to isolated communities in often harsh
511 environments. It also indicates that there are potential overlapping interests between local and
512 external institutions through the recognition of development areas such as transportation and
513 public infrastructure.
514 5.2 Institutional Mapping
515 <Insert Figure 4 about here>

14
516 Evaluating the institutional landscape visualized in Figure 4, it is clear there is a complex
517 interaction of stakeholders involved in disaster risk reduction in Sagarmatha National Park. The
518 main points of evaluation for this study are:

519 • Decision-making: The majority of the disaster response plan is guided by the vision of
520 national and global institutions. For example, the UNDP’s Sendai Disaster Framework
521 2015-2030, USAID’s Resilient Development Framework, Nepal’s NAPA continues to be a
522 major driver of local actions. While each framework highlights stakeholder participation,
523 the process is initiated with organizational objectives already outlined by the frameworks
524 being used. Because of this, the planning process that shapes the overall risk reduction
525 arena remains in the hands of external institutions. Therefore, while prioritization is often
526 dictated by local institutions, the knowledge received and mechanisms for
527 implementation are all dictated through external institutional guidelines. Additionally,
528 NGOs above can create conflicts in creating long-term sustainable decision-making
529 processes surrounding disaster risk reduction. For instance, several cases were reported
530 by Rounce et al. (37) in which government participation in program management
531 decreased due to the presence of NGO projects. Combined with the misrepresentation of
532 local voices in mountain communities, it is shown that NGO involvement is not
533 guaranteed to increase local participation in decision-making processes.
534
535 • Information Flows: We found that most of the information stays within the realm of the
536 international and national scales, with some information being extracted from the
537 regional level. Significant data collection goes towards generating reports and the
538 publications on issues such as governance, vulnerability, and disaster reduction
539 mechanisms. This data is used to influence policymakers at the national level in order to
540 strengthen the science-policy collaborations such as those seen in ICIMOD’s knowledge
541 platform that covers the Hindu Kush Himalaya region (47). This is backed by the
542 respondents of the household surveys, where 31% answered that information regarding
543 climate change was from static sources such as newspapers, television, and radio
544 broadcasts. This is supported through the logit model, which showed information
545 originating from NGOs and training have a higher coefficient (.242, .286) of positively
546 influencing risk perception than more static forms (.012, .076) such as newspaper, or
547 tv/radio. Therefore, more interactive flow of information between external and local
548 institutions can play a significant role in influencing how communities are perceiving GLOF
549 risk. Additionally, a flood warning remains predominately reactive as 30.4% of
550 participants rely on witnessing through sight and sound to know when flooding events are
551 occurring. With only 16.7% relying on sirens, and 2.2% from radio, it shows that
552 information dissemination between institutional scales remains largely disconnected.
553 Supporting this, the CART analysis showed that when respondents identified no sources of
554 warning, they also had almost no perception of risk involved (coefficient: 0.139) versus
555 having some form of warning system. With 26.1% of respondents identifying no warning

15
556 system, a more established communication networks must be maintained. It is also
557 important to note that women were shown to have less access to different forms of
558 information on climate change, which corresponds with lower risk perception compared
559 to men. As Nepal continues to move towards more localized approaches, it is a promising
560 sign knowledge dissemination will continue to increase. However as will be discussed in
561 the LAPA case study below (Figure 6), there are clear disconnects between knowledge
562 flow and adaptation actions.
563
564 • Funding Streams: Most funding mechanisms originate from international organizations
565 and are dispersed to the national government, and then disseminated through
566 government agencies (11, 12). While NGOs can offer a more decentralized network for
567 the distribution of program funds, inconsistencies in project funding remain a key barrier
568 to implementing risk reduction projects (35). Peniston et al. (33) for instance cites several
569 occasions in which LAPA projects never extend past planning periods from a lack of
570 established funding sources. Additionally, due to the reliance of NGOs that are external to
571 communities at risk, funded initiatives are not guaranteed to represent locally voiced
572 concerns (4). Therefore, while there are local planning processes taking place that identify
573 responsible funding parties (such as the DNPWC, among others) there is a delay in funding
574 dispersal as they disseminate from international and national levels. Coordination of
575 funds from independent sources will play a key role in creating more effective action and
576 reducing project delays that are currently inherent. Figure 4 highlights this, as most
577 adaptation funding comes from large-scale international organizations such as the UDNP,
578 GEF, and World Bank who disperse funding through programs such as the Least
579 Developed Country Fund (LDCF), Strategic Climate Fund (SCF) and the Nepal Climate
580 Change Support Program (NCCSP) (45, 52). Therefore, LAPA project funding originates
581 from the international scale and works through external institutional pathways until funds
582 are disseminated to villages. However, Regmi & Star (35) point out that dispersed LAPA
583 funds often fall well short of the necessary amount for successful implementation. This
584 case highlights the additional disconnect between external and local institutions through a
585 more traditional top-down approach to project development in which project funding can
586 be cumbersome and lead to challenges in effective local adaptation strategies.
587

588 6. Discussion
589 6.1 Agenda Setting
590 To carry out international agendas for climate change adaptation, the United Nations Climate
591 Change Commission and GEF group manage the Least Developed Countries Fund (LDCF). This
592 fund prioritizes the development of NAPAs for developing countries by these international
593 agencies who guide prioritization and provide financial funding for its implementation and is
594 highlighted in Figure 4 (45). Through this, prioritization often originates from international

16
595 institutions whose priorities are on national scales, rather than localized initiatives that
596 promote community-based adaptation measures. Shaw (39) highlights the need for the
597 consideration of local capacities and their roles in reducing localized risk in several cases
598 throughout Asia. Some of the more critical points mentioned are that strong local institutions
599 are vital to maintaining localized efforts and that these institutions should be integrated into
600 policies to ensure synergistic efforts. Gaventa (18) discusses how power in development
601 processes is multi-dimensional, expressed over at different scales, with varying degrees of
602 participation and visibility. This creates convoluted power structures in which development
603 actors often have disproportionate influence over decision-making processes, due to a lack of
604 transparency. The authors effectively convey how participatory projects at local levels are often
605 controlled by invisible actors at higher scales as seen below. Following Shaw (39), we argue that
606 a successful approach to adaptation requires new imagination that is guided by the vision of
607 synergy and transparency, different than commonly applied siloed, technocratic, bureaucratic
608 and top-down approach.
609 While Nepal’s NAPA is a national policy, it sets the structure of its more localized LAPA process.
610 Within this, influences of external priorities are witnessed such as the HiMAP which was a
611 partnership between USAID and NGOs. During this initiative, USAID proposed changes to the
612 LAPA process to better fit its Resilient Development Framework (6). There are also clear
613 indications of prioritization for project outcomes to fit within international development
614 objectives. Additionally, although Nepal’s government has initiated several LAPA projects
615 throughout the country, the rate of objective implementation remains low due to disjointed
616 organizational funding and regional organizational capacity (35).
617 Cases of the agenda-setting as highlighted above can have long term impacts on adaptation
618 planning efforts. Community trust in institutions plays a significant role in creating effective
619 GLOF mitigation measures (9). This can be seen in cases in Peru and Nepal in which institutional
620 insensitivity towards community values and priorities caused social resistance towards well-
621 intentioned mitigation strategies (13). Because of this, traditional and informal institutions
622 should be valued in creating community-based adaptation approaches that are more adept at
623 dealing with regional nuances in geography and culture that shape people’s perception towards
624 GLOF risk.
625 6.2 Institutional Disconnect
626 As shown above, there are several key barriers for institutions at different scales to address and
627 reduce the risk for GLOF events in the Solukhumbu region. One of the most significant barriers
628 is the misalignment of local and external institutional capacities and agendas. This is evident in
629 Nepal’s natural resource management history. Figure 5 shows how different factors govern the
630 risk reduction area. National policy promotes increasingly decentralized approaches to handling
631 the risk of GLOF events, yet they rarely interact with the indigenous institutions actively
632 managing disaster response measures (1). The disconnect between indigenous and national
633 institutions can lead to further conflict between stakeholders, where programs designed to
17
634 address inequity merely shifted power dynamics between groups. A historical example of the
635 replacement of traditional forest management systems with government-initiated community
636 user groups that reset traditional leadership roles through the creation of new social elites
637 within communities (49). Additionally, it is reported in LAPA processes that leadership roles
638 often exclude groups marginalized indigenous groups and women (33). This complemented by
639 discussed case studies from Peru highlights the importance of realizing the complex and often
640 nuanced systems of knowledge social structure that influence institutional management of
641 GLOF risk reduction.
642
643 <Insert Figure 5 about here>
644
645 Building on this finding, knowledge dissemination to local communities is lacking. Most external
646 institutions focus on the generation of knowledge for science-policy interactions. Therefore,
647 knowledge transfer remains within the external institutions, rather than to more isolated
648 mountain communities (29). Knowledge collected by organizations such as CFGORRP is largely
649 disseminated through static means such as pamphlets, signboards, or over the radio
650 transmission. This shows that knowledge pathways remain disjointed and noniterative, which
651 can increase confusion, fear, and discontent within communities. This is shown in the Tsho
652 Rolpa lake case study where the lack of transparency of government efforts to minimize risk
653 lead to increasing distrust towards external institutions (13). This is further supported in an
654 institutional capacity needs assessment conducted by CBFGORRP that analyzed each
655 department involved in climate adaption and disaster risk management. One of the major
656 recognized areas of need within each department is increased knowledge generation and its
657 communication to remote mountain regions. This reveals two major insights: (1) knowledge
658 dissemination from external to local institutions is inconsistent and largely lacking across
659 institutions; (2) Interorganizational cooperation is knowledge sharing is lacking clear pathways
660 across scales. This is evident as limited knowledge is listed as a weakness within each
661 department of the government responsible for some aspect of GLOF management. Additional
662 support for this claim comes from ICIMOD’s report on Disaster Risk Reduction in the Hindu Kush
663 Himalaya. In its key findings, the report emphasizes the importance of increasing the resiliency
664 of information systems that focuses on multi-stakeholder inclusivity (47).
665 The challenges described above highlight the legacy of institutional approaches to dealing with
666 Nepal’s mountain communities by the centralized government (34). Despite the rich tapestry of
667 knowledge on successfully dealing with local environmental stressors in the Khumbu region,
668 Sherpa communities still find themselves at odds with the central government contributing to
669 the policy process. This is captured by Wakiyama’s (49) analysis on the development of
670 community user groups that usurped traditional forest management systems over time. By
671 failing to recognize the value of informal institutions and causing power imbalances and discord
672 in mountain communities. Current external institutions often continue this trend in which
673 development projects are framed within their standards and objectives. Community input is

18
674 considered a part of each framework, yet it is not always practiced. For instance, the
675 Government of Nepal initiated risk reduction projects for Tsho Rolpa through artificial drainage
676 and Early Warning Systems (41). However, despite the emphasis placed on local participation in
677 their project framework communities were subjected to fear and confusion from the lack of
678 reporting by agencies responsible for the mitigation work (13).
679 This institutional gap is increasingly important as agency and organization for tourism efforts
680 manifest at local levels. For effective collaboration to take place within the Risk Reduction
681 Arena (Figure 5), external institutions must be able to effectively navigate the values placed on
682 tourism as an increasing source of livelihood (40). Tourism has brought Sherpa communities to
683 the global stage. With this, access to information is increasing, while institutional decision-
684 making and funding networks largely resemble historical mechanistic approaches often seen in
685 development projects (35). Examples can be found in Figure 4, where funding pathways
686 originate from the international level and the funds are disseminated through various
687 organizations until made available to VDCs who are managing LAPA projects (11). This can
688 create significant barriers for effective project implementation moving forward (35). To reduce
689 disconnect in the future, external funding agencies need to recognize the agency of local
690 institutions in managing adaptation funding in addition to local planning processes. This will be
691 a more long-term sustainability vision, in which communities play a prevailing role within the
692 management of decision processes, project funding, and information generation. External
693 Institutions must recognize the agency and priorities that are shaping Solukhumbu communities
694 as they move forward, and their capacity for managing adaptation strategies for unique SESs
695 that they are integral in shaping.
696 6.3 LAPA case study
697 Through the creation and prioritization of the LAPA process, some vital issues such as gender
698 equity and indigenous knowledge and inclusivity are now being built into institutional policy for
699 adaptation. However, there is still a disconnect between external and local institutions that is
700 revealed through LAPA efforts that were initiated by the UNDP in partnership with the
701 Government of Nepal from 2013-2017 shown in Figure 6 (11). Despite increasing objectives for
702 the inclusion of social equity in the planning process, major adaptation actions (such as the
703 prioritized drainage project of the Imja Lake) are often drastically delayed in their
704 implementation. For instance, construction planning did not begin until 2016, leaving less than
705 a year of the project duration for adaptation implementation due to the bureaucracy of the
706 construction contract bidding process (12). This disconnect between knowledge being
707 disseminated to community members and concrete actions being taken can exacerbate existing
708 mistrust between communities and external institutions. Institutional gaps such as this highlight
709 the need for an effective policy that addresses.
710 <Insert Figure 6 about here>

19
711 This provides a concrete example of how the institutional gaps in Figure 4 regarding financial
712 flows and knowledge dissemination can cause potential conflict in the Risk Reduction Arena in
713 Figure 5. This disconnect between institutions regarding knowledge flow and adaptation actions
714 can result in the impediments such as perceived risk, knowledge disparity, and institutional
715 instability when project cycle transparency becomes an issue (8).
716 7. Conclusions
717 Overall, GLOF is an increasingly threatening hazard and risk in the lives of mountain
718 communities in the SNPBZ and downstream communities. People’s perception of risk, and
719 therefore, their response to adaptation measures is influenced by a complex mix of social
720 factors such as livelihoods, exposure, and spirituality that all shape an individual’s evaluation of
721 his or her vulnerability to glacial hazards. While these communities have a strong history of
722 creating local adaptation strategies that have shaped the geography and their way of life,
723 tumultuous history of the national government has changed the institutional landscape and
724 communities’ relationship to the resources they rely on (49). Shifting power dynamics towards
725 a centralized approach to research management saw the characterization of mountain
726 communities as periphery to the overall interest of the national government. This attitude
727 changed the dichotomy of resource management from one of localized participation and use
728 towards a more exploitive regional development strategy, including the establishment of
729 national parks in these mountains and the promotion of tourism industry as a rent-seeking
730 mechanism (34). This period saw the introduction of an international agenda-setting, as
731 international aid agencies influenced national policies towards a more pro-development
732 agenda and economic liberalization. More recent efforts to create decentralized management
733 strategies have led to the disjointed creation of formal local institutions such as buffer zone
734 user groups that often exist in direct competition to the institutions (49). This condition shows
735 that there is a clear disconnect between external institutions and local informal institutions that
736 have been shaped by unique geography, culture, and social structure. As argued by Klein et al
737 (24), high mountains like the Himalayas that are highly biodiverse and have multi-hazard zones
738 need a specialized disaster risk reduction strategy, which considers both biophysical and socio-
739 cultural realities of the mountains. In this paper, we illustrated the institutional dynamics as
740 well as the processes involved in GLOF hazards and risks mitigation projects in the Himalayas
741 that are largely driven by external agenda, technological fix model, and bureaucratic exercises
742 to the extent that local communities feel disconnected from the processes.
743 As community-based disaster risk reduction becomes more of a national priority (such as in
744 Nepal’s NAPA), the value of local and informal institutions needs to be recognized by external
745 institutions and play a more critical role in developing adaptation strategies to glacial hazards.
746 This approach entails not only the incorporation of local knowledge into risk reduction
747 strategies at the local level but also the recognition of informal decision-making structures into
748 national policy. Several additional conditions should be considered for effective institutional
749 action to take place. Knowledge dissemination needs to be developed into clear pathways that

20
750 address different organizations at each scale. Lack of consistency and coherence in the
751 communication of scientific knowledge and findings can be a clear and significant barrier to
752 effective inter-organizational coordination in disaster risk management. This applies in the case
753 of Nepal too, as there were several gaps and disconnects between the Government of Nepal,
754 national and international organizations, and the incorporation of local knowledge into policy.
755 In addition to the coordination of information, funding streams need to be effectively managed
756 between institutions to ensure project continuity and scalability.
757 While the institution of GLOF management has changed from external agencies to a more
758 localized process, it is important to recognize that these organizations have shaped this process
759 and are still present. The implementation of LAPA processes marks an important milestone in
760 external institutions’ recognition of localized adaptation strategies through building community
761 agency and self-organization as described by Shaw (39). However, the funding mechanisms
762 remain disconnected and delayed from the LAPA process, barring the adaptation strategies
763 prioritized from becoming a reality (35). Moving forward, governance of disaster risk should
764 ensure effective governance of local institutions that have both social credibility and legitimacy
765 among the residents. As shown by Van Den Berg (48), multi-scalar governance is possible when
766 a national government proactively works to strike a balance between engineering approach
767 (e.g., dikes, barriers) and multilevel adaptation strategies (e.g., spatial planning, crop rotation,
768 water boards, relocation of freshwater intake points) in the Netherlands—a country below the
769 sea level facing perennial flooding risks. This multi-scalar adaptation approach can help to
770 overcome the traditional difficulties of centralized strategies attempting to implement projects
771 under the difficulties of difficult terrains and disconnected knowledge of local concerns. Overall,
772 this paper has painted an institutional landscape that visualizes information, policy, and funding
773 networks and connects them to a more theoretical understanding of community-based
774 adaptation efforts. Additionally, it forms a connection to the perceptions at the household level
775 that offers key insight into how community members perceive adaptation efforts moving
776 forward.
777

778

779

780

21
781 Acknowledgements
782 This material is based upon work supported by the National Science Foundation under Grant
783 Number ICER-1516912, CNH-L: Science-Driven, Community-Based Approach to Reducing Glacier
784 Lake Outburst Flood Risks--a multidisciplinary research collaboration of University of Texas
785 Austin (PI: D Mckinney), Arizona State University (Co-PI: M Shrestha), and University of
786 Colorado, Boulder (Co-PI: A Byers). This research benefitted greatly from the social survey
787 dataset created by this collaborative project. We would like to thank local leaders, research
788 participants and respondents for welcoming us and giving us valuable time. J Parajuli, S
789 Adhikari, SF Sherpa, S Sharma, LM Wagle, S Paudyal and other assisted in research fieldwork
790 and data collection; their assistance is appreciated. We would like to thank the three
791 anonymous reviewers for their constructive feedback and insights which helped improve this
792 paper. The authors, however, are solely responsible for any error and omission, and the
793 findings and views reflected here are those of the authors, not the funding agency.
794

795

796

797

798

799

800

801

802

803

804

805

806

807

808

809

810

22
References

1) A. Agrawal and C.C. Gibson. The Role of Community in Natural Resource Conservation
Communities and the Environment: Ethnicity, Gender, and the State in Community-Based
Conservation. London: Rutgers University Press, 2001.
2) I.-M. Bjønness. Mountain Hazard Perception and Risk-Avoiding Strategies among the
Sherpas of Khumbu Himal, Nepal. Mt. Res Dev. 6 (1986) 277–292.
3) F. Berkes, H. Ross. Community Resilience: Toward an Integrated Approach, Society &
Natural Resources, 26 (1) (2013) 5-20, DOI:10.1080/08941920.2012.736605
4) D. Bishwokarma, N. S. Paudel, J. J. Molina. Civil Society Organizations in Climate Change
Policy Dynamics in Nepal: How Representative and Inclusive is the Process? TMC Academic
Journal, 7(1) (2012), 1–12.
5) A.C. Byers AC, A. D. Cuellar, D. C. McKinney. High Mountains Adaptation Partnership:
Lessons Learned in Nepal and Peru. Washington D.C.: USAID, 2015.
6) A. C. Byers, S. Thakali. Khumbu Local Adaptation Plan of Action (LAPA): Sagarmatha
National Park, Solukhumbu District. Washington D. C.: USAID, 2016.
7) M. Carey. In the Shadow of Melting Glaciers: Climate Change and Andean society. Oxford
University Press, 2010.
8) M. Carey, C. Huggel, J. Bury, C. Portocarrero, W. Haeberli, An integrated socio-
environmental framework for glacier hazard management and climate change adaptation:
lessons from Lake 513, Cordillera Blanca, Peru, Clim. Change. 112 (2012) 733–767.
https://doi.org/10.1007/s10584-011-0249-8.
9) M. Carey, G. McDowell, C. Huggel, J. Jackson, C. Portocarrero, J.M. Reynolds, L. Vicuña,
Integrated Approaches to Adaptation and Disaster Risk Reduction in Dynamic Socio-
cryospheric Systems. In W. Haeberli, C. Whiteman (eds) Snow and Ice-Related Hazards,
Risks and Disasters, Elsevier, 2015. https://doi.org/10.1016/B978-0-12-394849-6.00008-1.
10) J.L. Carrivick, F.S. Tweed, A global assessment of the societal impacts of glacier outburst
floods, Glob. Planet. Change. 144 (2016) 1–16.
https://doi.org/10.1016/j.gloplacha.2016.07.001.
11) CFGORRP. Institutional Capacity Needs Assessment. Community Based Flood and Glacial
Lake Outburst Risk Reduction Project (CFGPRRP). Kumari Kripa Marg, Samakhusi,
Kathmandu, 2014.
12) CFGORRP. What we achieved so far. United Nations Development Program (UNDP). (2017)
13) K.R. Dahal, R. Hagelman, People’s risk perception of glacial lake outburst flooding: a case
of Tsho Rolpa Lake, Nepal, Environ. Hazards. 10 (2011) 154–170.
https://doi.org/10.1080/17477891.2011.582310.
14) P.J. DiMaggio, W.W. Powell, The Iron Cage Revisited: Institutional Isomorphism and
Collective Rationality in Organizational Fields, Am. Sociol. Rev. (1983).
https://doi.org/10.2307/2095101.
15) DNPWC. Sagarmatha National Park Management and Tourism Plan 2006–2011.
Department of National Parks and Wildlife Conservation (DNPWC), Kathmandu, Nepal,
2006.

23
16) DNPWC. Sagarmartha National Park and its Buffer Zone Management Plan 2016-2020.
Department of National Parks and Wildlife Conservation (DNPWC), Namche Bazaar,
Solukhumbu, Nepal, 2016.
17) K. Engel, G. Frerks, L. Velotti, J. Warner, B. Weijs, Flood disaster subcultures in The
Netherlands: The parishes of Borgharen and Itteren, Nat. Hazards. (2014).
https://doi.org/10.1007/s11069-014-1116-5.
18) J. Gaventa, Finding the spaces for change: A power analysis, IDS Bull. (2006).
https://doi.org/10.1111/j.1759-5436.2006.tb00320.x.
19) ICIMOD, Glacial lakes and glacial lake outburst floods in Nepal. International Centre for
Integrated Mountain Development (ICIMOD), Kathmandu, Nepal, 2011.
20) IPCC, Summary for Policymakers. In: IPCC Special Report on the Ocean and Cryosphere in a
Changing Climate., 2019. https://doi.org/https://www.ipcc.ch/report/srocc/.
21) R. Karkee, J. Comfort, NGOs, Foreign Aid, and Development in Nepal, Front. Public Heal.
(2016). https://doi.org/10.3389/fpubh.2016.00177.
22) R.W. Kates, W.C. Clark, R. Corell, J.M. Hall, C.C. Jaeger, I. Lowe, J.J. McCarthy, H.J.
Schellnhuber, B. Bolin, N.M. Dickson, S. Faucheux, G.C. Gallopin, A. Grübler, B. Huntley, J.
Jäger, N.S. Jodha, R.E. Kasperson, A. Mabogunje, P. Matson, H. Mooney, B. Moore, T.
O’Riordan, U. Svedin, Environment and development: Sustainability science, Science (80-.
86). (2001). https://doi.org/10.1126/science.1059386.
23) J. B. Khatri-Chhetri. Indigenous management of Forest Resources: A Case Study of Jomson.
VDC in Mustang District, From Indigenous Management of Natural Resources in Nepal. In
D. Tamang, G.J Gill and C.B. Thapa (ed.) Indigenous Management of Natural Resources in
Nepal. Winrock Institutional Policy Analysis in agriculture and related resource
management, 1993.
24) J.A. Klein, C.M. Tucker, C.E. Steger, A. Nolin, R. Reid, K.A. Hopping, E.T. Yeh, M.S. Pradhan,
A. Taber, D. Molden, R. Ghate, D. Choudhury, I. Alcántara-Ayala, S. Lavorel, B. Müller, A.
Grêt-Regamey, R.B. Boone, P. Bourgeron, E. Castellanos, X. Chen, S. Dong, M. Keiler, R.
Seidl, J. Thorn, K. Yager, An integrated community and ecosystem-based approach to
disaster risk reduction in mountain systems, Environ. Sci. Policy. 94 (2019) 143–152.
https://doi.org/10.1016/j.envsci.2018.12.034.
25) G. McDowell, C. Huggel, H Frey, F. M. Wang, K. Cramer, V. Ricciardi. Adaptation action and
research in glaciated mountain systems: Are they enough to meet the challenge of climate
change? Global Environmental Change, 54, (2019) 19–30
26) Ministry of Environment. National Adaptation Programme of Action (NAPA) (2010).
Kathmandu, Nepal, 2010.
27) J. V. Murra. An Aymara Kingdom in 1567. Ethnohistory, 15(2) (1968)115-151.
28) National Planning Commission. Sustainable Development Goals, 2016-2030, National
(Preliminary) Report. Government of Nepal, National Planning Commission, Kathmandu,
Nepal, 2015.
29) H.R. Ojha, R. Ghate, L. Dorji, A. Shrestha, D. Paudel, A. Nightingale, K. Shrestha, M.A.
Watto, R. Kotru, Governance: Key for Environmental Sustainability in the Hindu Kush
Himalaya, in: Hindu Kush Himalaya Assess., Springer International Publishing, 2019: pp.
545–578. https://doi.org/10.1007/978-3-319-92288-1_16.

24
30) E. Ostrom, Governing the commons: the evolution of institutions for collective action,
Cambridge University Press, Cambridge England; New York, 1991.
31) E. Ostrom, Polycentric systems for coping with collective action and global environmental
change, Glob. Environ. Chang. 20 (2010) 550–557.
https://doi.org/10.1016/j.gloenvcha.2010.07.004.
32) E. Ostrom, Background on the Institutional Analysis and Development Framework, Policy
Stud. J. 39 (2011) 7–27. https://doi.org/10.1111/j.1541-0072.2010.00394.x.
33) B. Peniston. High Mountains Adaptation Partnership: A Review of Nepal’s Local Adaptation
Plans of Action (LAPA). Kathmandu, Nepal: The Mountain Institute, 2013.
34) R. E. Rhoades. Listening to the Mountains. Iowa: Kendall/Hunt, 2007.
35) R. Shaw, Community-Based Disaster Risk Reduction, 1 (2016) 1–24.
https://doi.org/10.1093/acrefore/9780199389407.013.47.
36) B.R. Regmi, C. Star, Identifying operational mechanisms for mainstreaming community-
based adaptation in Nepal, Clim. Dev. (2014).
https://doi.org/10.1080/17565529.2014.977760.
37) D.R. Rounce, D.C. McKinney, J.M. Lala, A.C. Byers, C.S. Watson, A New Remote Hazard and
Risk Assessment Framework for Glacial Lakes in the Nepal Himalaya, Hydrol. Earth Syst.
Sci. Discuss. (2016) 1–48. https://doi.org/10.5194/hess-2016-161.
38) D.R. Rounce, D.J. Quincey, D.C. McKinney, Debris-covered glacier energy balance model for
Imja–Lhotse Shar Glacier in the Everest region of Nepal, Cryosph. 9 (2015) 2295–2310.
https://doi.org/10.5194/tc-9-2295-2015.
39) W. Schwanghart, R. Worni, C. Huggel, M. Stoffel, O. Korup, Uncertainty in the Himalayan
energy-water nexus: Estimating regional exposure to glacial lake outburst floods, Environ.
Res. Lett. (2016). https://doi.org/10.1088/1748-9326/11/7/074005.
40) S.F. Sherpa, M. Shrestha, H. Eakin, C.G. Boone, Cryospheric hazards and risk perceptions in
the Sagarmatha (Mt. Everest) National Park and Buffer Zone, Nepal, Nat. Hazards. (2019).
https://doi.org/10.1007/s11069-018-3560-0.
41) B.B. Shrestha, H. Nakagawa, Assessment of potential outburst floods from the Tsho Rolpa
glacial lake in Nepal, Nat. Hazards. 71 (2014) 913–936. https://doi.org/10.1007/s11069-
013-0940-3.
42) M.A. Somos-Valenzuela, D.C. McKinney, A.C. Byers, D.R. Rounce, C. Portocarrero, D.
Lamsal, assessing downstream flood impacts due to a potential GLOF from Imja Tsho in
Nepal., Hydrol. Earth Syst. Sci. 19 (2015) 1401–1412. 10.5194/hess-19-1401-2015.
43) S. F. Stevens. Claiming the high ground: Sherpas, subsistence, and environmental change in
the highest Himalaya. Motilal Banarsidaas Publisher, India, 1996.
44) T. Terpstra. Emotions, trust, and perceived risk: Affective and cognitive routes to flood
preparedness behavior. Risk Analysis: An International Journal, 31(10) (2011) 1658-1675
45) UNDP. Community Based Flood and Glacial Lake Outburst Risk Reduction Project
(CFGORRP), United Nations Development Programme (UNDP), Nepal, 2014.
46) UNISDR, 2015. Desinventar Disaster Information System.
https://www.desinventar.net/DesInventar/main.jsp (last visited May 2019)
47) Vaidya, A., Shrestha, M. S., & Nasab, N. Disaster Risk Reduction and Building Resilience in
the Hindu Kush Himalaya. In the Hindu Kush Himalaya Assessment (2019) 389–419.
Springer.

25
48) M.M. Van Den Berg, Climate Change Adaptation in Dutch Local Communities Risk
Perception, Institutional Capacity and the Role of Local Government, n.d.
http://www.utwente.nl/cstm. (accessed January 10, 2020).
49) T. Wakiyama, Community Forestry in Nepal: A Comparison of Management Systems
between Indigenous Forestry and Modern Community Forestry, 2004.
http://www.iges.or.jp/en/fc/pdf/report11/full_text.pdf (accessed February 24, 2019).
50) T. Watanabe, D. Lamsal, J.D. Ives, Evaluating the growth characteristics of a glacial lake and
its degree of danger of outburst flooding: Imja Glacier, Khumbu Himal, Nepal, Nor. Geogr.
Tidsskr. - Nor. J. Geogr. 63 (2009) 255–267. https://doi.org/10.1080/00291950903368367.
51) M.J. Westoby, N.F. Glasser, J. Brasington, M.J. Hambrey, D.J. Quincey, J.M. Reynolds,
Modelling outburst floods from moraine-dammed glacial lakes, Earth-Science Rev. 134
(2014) 137–159. https://doi.org/10.1016/j.earscirev.2014.03.009.
52) World Bank. Building Resilience to Climate- Related Hazards. Project Appraisal Document.
Washington D.C., 2012.

26
Figure 1. Study Site (map by Milan Shrestha)
Figure 2. Sagarmatha National Park Buffer Zone Structure (adapted from DNPWC 2016)
Figure 3. Classification and Regression Tree Analysis of Risk Perception based on GLOF warning
Figure 4. Institutional Map for Disaster Risk Management in Nepal

Solid lines represent current documented flows of disaster risk information. Dotted lines represent
financial flows, indicating a top-down funding pathway from international agencies to local levels for
LAPA plans.
Figure 5. Institutional conditions shaping the risk reduction arena.

The underlying conditions influencing institutional actions at different scales. (1) External institutions are
influenced by frameworks and policies that lead to knowledge dissemination, policy, and capacity
building for risk reduction. (2) Local institutions operate under social factors that create agency and self-
organization influencing their decisions within the risk reduction arena (Berkes & Ross, 2013). (3) SES
factors surrounding GLOF management also influence the risk reduction arena (Carey et al., 2012).
Figure 6. Timeline of the Imja Lake Lowering Project by CFGORRP/UNDP
Appendix A. 2016 – 2017 questionnaire subset

Question
What type of the problems you have encountered in water supply?
What was your response to cope with the events?
How would you know if the flooding is to occur?
What is the source from which you heard of climate change? Check all that apply.
If not, what is your main livelihood or income source?
In your opinion, what are the main challenges or problems for tourism in Khumbu?
What is the major natural disasters or uncertainties do you consider significant for tourism?
Choose one
Would you relocate your business to different location, if the threat of flooding is
imminent?
If you are concerned about the lake, who do you think should take the lead to do
something?
What do you think should pay for draining?