Money and knowledge have understandably figured prominently in the post-disaster discussion. Money helps us rebuild. Knowledge helps us understand. The harder challenge is turning both into the capability to act, learn, and do better the next time. This essay focuses on that third leg: capability.

The Debate So Far

Debates following recent Himalayan disasters have tended to emphasize three important dimensions:

Climate justice: Nepal contributes little to global emissions (less than 0.1%) yet bears growing climate-related risks, strengthening the case for Loss and Damage and international climate support.

Science and early warning: satellites, hydrological monitoring, forecasting and increasingly artificial intelligence can help us better understand and anticipate Himalayan hazards.

Governance and development choices: roads, hydropower, settlement patterns, river extraction, building practices and weak enforcement can create or amplify exposure to those hazards.

Each matters. But this essay asks a somewhat different question:

How do we turn what we know into what we can do together?

I think of this as distributed capability.

The idea has suddenly become more concrete. In his recent address to the UN General Assembly, Prime Minister Balendra Shah proposed a Himalayan Climate Resilience Mechanism led by Nepal, India and China, including shared satellite data and expertise, monitoring of dangerous glacial lakes, joint early-warning systems and coordinated cross-border disaster response.

That would be an important step. But seeing more does not automatically mean doing better. The proposal connects science across borders. This essay asks the next question: how does that science become capability across levels—from satellites and scientists to governments, municipalities and communities on the ground?

In essence, knowing is not enough. Scientific information must become practical know-how.

Taken together, these challenges point toward what I call a Geography of Learning. In the pages that follow, I try to explore what such a learning system might look like.

Exploring the Mustang-Lumbini Corridor

More than a decade ago, in December 2014, I traveled with a small interdisciplinary team from the University of New Mexico along what I had begun thinking of as the Gandaki River Corridor—from the high Himalaya of Mustang through the middle hills and toward the plains, with Lumbini as the final leg of our journey. I was joined by my colleagues Joe Galewsky, an atmospheric scientist, and Mark Stone, an engineer, together with several graduate students and an environmental scientist, Dr. Siddhartha Bajracharya of NTNC.

We were traveling with a modest NSF seed grant and a fairly exploratory question: could this extraordinary stretch of Nepal, running from the high Himalaya toward the plains, become something of a living laboratory?

In Marpha, conversations quickly brought the abstraction of climate change down to earth. Local people spoke about changes they were already observing—less snowfall, unusual rainfall patterns, and consequences for apple farming. They even spoke of snow leopards being sighted around their alpine cattle-grazing areas, something they associated with a receding snowline.

Joe was interested in changing snowfall, aerosols and particulate matter in the atmosphere, and the Himalayan climate; Mark approached the landscape through rivers, hydrology, infrastructure and ecological resilience.

I kept thinking about how communities and schools might become participants in observing these changes through citizen science, inspired by an initiative that had begun in 1996 as an NSF-funded project out of UNM’s Biology Department under Dr. Clifford Crawford. What started modestly eventually grew into a full-fledged program led by a local school, involving dozens of schools, currently supported by state funding, and inspiring similar efforts in other river systems around the country. It is known as the Bosque Ecosystem Monitoring Program (BEMP).

In the years that followed, I continued exploring some of these ideas through research and experiential learning. We even tried a small BEMP-style citizen-science project with three schools along the Danda River in Siddharthanagar, where students participated in ecological monitoring and data collection. The experiment lasted only about six months, until the seed funding ran out, but it offered a small glimpse of what connecting schools, communities and scientific observation might look like.

There was another dimension to the experiment as well. We brought doctoral research and data collected from different districts of Lumbini Province into the undergraduate classroom at UNM, where students worked with the findings and thought about how they might be translated into something useful on the ground. Some of those ideas then traveled back to Nepal as small, actionable initiatives—from the Danda River monitoring project and community health workshops to installing air-pollution monitoring devices in schools across the district.

It was a modest form of experiential learning: university researchers, undergraduate students, schools and community leaders learning from one another, with knowledge moving back and forth between research, classroom and community. That hands-on model became part of the inspiration for the larger corridor idea.

During a UNM visit more than a decade ago, Kathmandu University’s founding vice chancellor, Dr. Suresh Raj Sharma, who recently passed away, visited the Bosque School and observed the BEMP citizen-science model firsthand. He and his team immediately saw the possibilities of connecting university research, a network of local schools along the river system, and communities through this kind of citizen science. The difficulty was not seeing the idea; it was making room for it within limited extracurricular funding and an education system that remained largely prescriptive, leaving little space for exploratory experiments of this kind.

This was also when the Mustang–Lumbini Corridor began to emerge. I returned to it several times over the years, including in Nepal Unplugged essays about a Mustang-to-Lumbini science corridor and what I somewhat ambitiously called a “Vertical University.” At UNM, I also developed a Himalayan study-abroad program in partnership with a local college in Siddharthanagar, blending STEM with community-engaged research, awareness and learning.

Looking back now, however, I think the more interesting question was not about the particular program I was imagining—whether our little experiment survived, or why Kathmandu University could not assemble schools and communities into a data-sharing network like BEMP.

Nepal has no shortage of such small experiments. The harder challenge is whether they end with the grant and the project report, or leave behind relationships, routines and capabilities that continue to learn—so that we are better prepared when the next catastrophic shock arrives.

In essence, the question underneath is this:

How do we turn what we know into what we can do together?

That question feels considerably more urgent today.

A Decade of Shocks

The years since that Mustang journey have confronted Nepal with repeated and very different shocks.

The 2015 earthquake was followed by major monsoon flooding in 2017, the devastating Melamchi flood in 2021, the floods and landslides of 2024, and now another catastrophic Himalayan event in 2026.

Each disaster is different. We should be careful about collapsing earthquakes, monsoon floods, landslides, glacial hazards and infrastructure failures into a single causal story. But together they raise an uncomfortable question.

Can Nepal turn repeated shocks into cumulative learning?

This is somewhat different from the questions I explored in the first two essays of this series.

Part I asked us to reconsider the Geography of Development—particularly the interconnected mountain and river systems that frequently cut across our administrative boundaries. More broadly, I argued that Nepal’s administrative federalism might be complemented by what I called developmental federalism: looking across political boundaries to recognize the ecological and economic systems that connect regions.

Part II traced what we might call the Geography of Risk: how development choices create different forms of exposure—through engineering decisions, investment choices and locational preferences, alongside the growing risks associated with global climate change—and how shocks can destroy not only infrastructure but health, education, livelihoods, social capital and future capabilities.

But another question remains.

Once scientific studies wrap up, official reports are published, and public interest begins to fade, political and policy attention often moves on. Some lessons take root; others fade with projects, personnel and institutional attention—until the next crisis brings the same questions back again.

So, what have we learned?

Knowing More Is Not Knowing How

Nepal certainly needs better information.

Important gaps remain in glacier monitoring, hydrology, slope instability, local hazard mapping and early warning. The rapidly changing Himalayan environment will make better observation even more important.

But lack of information cannot be the entire explanation.

Nepal also possesses considerable scientific capacity. Universities, government technical agencies, the International Centre for Integrated Mountain Development (ICIMOD), international researchers and communities themselves have accumulated substantial knowledge about Himalayan hazards.

Some of the connections imagined here exist as promising prototypes. SERVIR-Hindu Kush Himalaya, an earlier NASA-USAID initiative implemented through ICIMOD, used satellite Earth observation and geospatial tools for applications ranging from water and land management to disaster risk. Its wonderfully simple formulation—“connecting space to village”—captures much of the challenge. In 2019, the initiative trained high-school teachers from rural Nepal in Earth-observation and geospatial tools, attempting to link satellite data to school classrooms. Nepal, in other words, is not starting from zero.

Nor has Nepal lacked policy ambition. The Disaster Risk Reduction and Management (DRRM) Act of 2017 marked an important shift toward a proactive risk-reduction framework, while the updated National Building Code, including NBC 105:2020, strengthened the formal mechanism for safer construction.

But laws, codes, monitoring systems and scientific project interventions bring us back to the same question:

When does formal knowledge become routine practice?

The recent discussion surrounding the Rapid Damage and Needs Assessment (RDNA) following the Rasuwa–Bhote Koshi disaster brings this problem into sharp relief. In a thoughtful critique of the assessment, Apil KC asks a deceptively simple question: what is such a document actually supposed to enable? Nepal has accumulated considerable experience from earlier disasters, including the Gorkha earthquake and Jajarkot. The deeper test, therefore, is not simply whether each disaster produces a better assessment, but whether what is learned leaves institutions better able to act before and after the next one.

My recent reading around information, complexity and networks has made me look at this distinction somewhat differently.

Information and know-how are not the same thing.

A hazard map contains information. A scientific paper contains information. A satellite can produce extraordinary amounts of information.

But knowing how to use that map when approving a building, redesigning a hydropower intake, evacuating a village or deciding where not to build requires something else. The same is true of risk. As I discussed in Part II, understanding different forms of risk matters only if that understanding actually enters decisions about infrastructure, settlements and preparedness.

This is where knowledge has to become know-how.

Know-how resides partly in people. More importantly, increasingly complex know-how resides between people—in relationships, complementary skills, repeatable routines, organizations and networks that perform a task together under stress. A flood warning, for example, saves lives only when forecasters, local officials, communities and families know how to act together.

Nepal’s challenge, then, is not simply to produce more knowledge about disasters, but to turn that knowledge into something people and institutions can actually use, together and repeatedly. So, while a trilateral Himalayan monitoring and information-sharing mechanism involving China, Nepal and India could be an important step, the harder challenge lies beyond it: converting that information into actionable know-how on the ground, across the landscape.

Evidence That Learning Can Work

It would be wrong, however, to tell this as another story of a country that never learns.

Nepal’s experience with community-based flood early-warning systems points in a more hopeful—and analytically interesting—direction.

Nepal’s Department of Hydrology and Meteorology introduced its community-based Flood Forecasting and Early Warning Services in 2016 and subsequently expanded warnings through SMS, social media, radio, television and email. The World Meteorological Organization reports that national flood deaths fell from 166 in 2017 to 27 in 2022, even though it characterized the 2022 flooding as being of higher magnitude.

That comparison does not establish that early warning alone caused the decline—weather patterns, flood timing and exposure vary. But it is consistent with what we would expect when forecasting becomes connected to communication, preparedness and practiced local response.

Gauge readings alone cannot evacuate a village. Neither can an automated SMS message. There are accounts from the recent disaster of people downstream receiving warnings with time to evacuate, yet not immediately acting on them.

Thus, for information to become safety, somebody has to interpret the warning. Somebody has to know whom to contact. Local disaster-management committees must mobilize. Families need to understand what the signal means. Routes and safe places must be known in advance. The relationships have to work when the river rises.

Where these physical and social links have been repeatedly connected and practiced, information generated from above has begun to become capability on the ground.

Melamchi offers another kind of lesson. ICIMOD’s assessment of the devastating 2021 disaster found no single cause. Heavy rainfall and rapid snowmelt interacted with glacial deposits, landslides, river blockage, erosion and massive sediment movement. One hazard amplified another as the disaster moved downstream.

The same river system had also been heavily mined for sand and gravel to feed Kathmandu’s construction appetite—buildings, housing colonies and other urban development. Streams of Kathmandu-bound tippers created their own environmental and health burdens for communities along the route in Bahunipati.

I had seen this firsthand. The summer after the 2015 earthquake, I spent about a month there with UNM4Nepal, a group largely made up of UNM engineering students, working in collaboration with Kathmandu University to build a women’s community center at KU’s Bahunipati clinic. Mark Stone and I served as faculty mentors. I still remember children playing along a road constantly clouded with dust from the stream of gravel trucks. I complained to Kathmandu University and encouraged the local women’s group to raise the issue as well.

Development choices, river systems and local lives were already intertwined.

Much was learned after Melamchi. But the real measure of learning is whether those lessons change what we do next—where and how we build, how we manage rivers, and how we prepare communities downstream.

In a sense, the Melamchi story is where all three geographies meet: development, risk and learning. How we develop shapes the risks we create; what we learn from those risks should shape how we develop next.

This brings us back to the proposed Himalayan mechanism involving Nepal, India and China. Shared satellite data, monitoring and new AI tools could greatly improve what we know about these changing mountain systems. But information from above still has to meet realities on the ground. A satellite may detect a dangerous change upstream; somebody downstream still has to understand the warning, trust it and know what to do.

That is the missing connection: science from above must become capability on the ground. And that capability must help us make better development choices, understand the risks they create, and learn before the next disaster arrives.

From Above and On the Ground

Our capacity to see the Himalaya from above is expanding rapidly.

Satellites can observe glaciers and snow cover. Remote sensing can identify landscape changes. Hydrological models can track rivers. Artificial intelligence may increasingly help detect patterns that human observers cannot easily see.

The Mekong offers one useful precedent for sharing such information across borders. Cambodia, Lao PDR, Thailand and Viet Nam cooperate through the Mekong River Commission, sharing basin information and coordinating monitoring, forecasting and water management across national boundaries. China and Myanmar, the two upstream countries, participate as Dialogue Partners. The Himalayan political geography is different, but the underlying principle is relevant: a river system does not stop at a national border.

The Himalayan “Third Pole” similarly connects China, Nepal and India through shared mountain and river systems. Prime Minister Shah’s proposed Himalayan Climate Resilience Mechanism now puts the possibility of trilateral scientific monitoring and information sharing more concretely on the table.

But sharing information across borders is not always easy. Questions of trust, security and national interest can get in the way, even when the science itself is clear.

And seeing more does not automatically mean doing better.

Government also has to remember what worked and what did not, so that lessons survive after officials and governments change. This is what I have often referred to as institutional memory.

Information should not travel in only one direction—from scientists and government down to communities. People living in these landscapes observe things too.

During an NDRI-organized panel discussion, former UN Ambassador Dr. Jaya Raj Acharya recalled a Sherpa drawing his attention to an unusual change in high-altitude water flow. Whether such an observation signals something scientifically important is a question for further investigation. But the story raises a very practical question:

Where does such locally observed information go?

There are already simple examples of how this can work. Cornell University’s eBird allows ordinary birdwatchers around the world to report what they see, helping scientists understand changes in bird distribution and migration. CrowdWater, developed at the University of Zurich, does something similar with water: citizens contribute observations of water levels, stream conditions and soil moisture.

Imagine something similar across Nepal’s river basins. Local people could contribute what they observe, while scientists and government agencies could share what satellites, monitoring stations and models are showing.

Information would then travel both ways—from satellite to village, and from village back to scientists and decision-makers.

This is also where regional universities could play an important role.

Rather than functioning simply as institutions located in different parts of Nepal, parts of their teaching and research could become more deeply embedded in the ecology and landscape around them. Universities and research centers across the high Himalaya, middle hills and Terai could develop complementary capabilities around glaciers and snow, watersheds and landslides, forests, rivers, agriculture, groundwater, flooding and other locally relevant systems.

They need not control the network. They could become regional knowledge anchors—connecting schools, communities, scientists, local governments and national institutions, while helping preserve what is learned from one generation of students and researchers to the next.

This could be especially useful for smaller municipalities. We cannot expect every municipality to employ its own hydrologists, geologists, climate scientists and other specialists. But several municipalities could draw upon expertise available through a nearby university or research center. This is one simple example of distributed capability: the knowledge and skills do not have to reside in one place to be available to the larger system.

Local governments themselves must be part of this learning process. The institutions closest to the ground cannot learn from experience if they are left outside important decisions.

So the challenge is not simply to collect more information. It is to connect people and institutions so that what one part of the system learns can be shared, remembered and used by others.

This brings me back to Mustang. What villagers in Marpha knew from living within their landscape was different from what an atmospheric scientist could observe through instruments or what an engineer could calculate.

Neither form of knowledge substitutes for the other.

The opportunity lies in connecting them.

Imagine the Himalayan river basin not simply as a geography through which water, sediment, people, roads, electricity and commerce move, but as a Geography of Learning.

The Geography of Learning

So what do I mean by a Geography of Learning?

Nepal’s rivers connect mountains, valleys and plains. A flood or landslide in one place can have consequences far downstream. Learning must be able to travel across that physical landscape too.

If Melamchi teaches us something about cascading hazards, that lesson should influence decisions in other river basins. If one community learns how to respond effectively to an early warning, other communities should be able to learn from that experience. And those lessons should survive changes in projects, municipal staff and political administrations.

A Geography of Learning, as I see it, brings together four parts:

Science from Above. Satellites, hydrological monitoring, scientific models and increasingly artificial intelligence can give us a wider view of changing conditions across river basins, including those that cross national borders.

Capability on the Ground. Citizens, schools and local governments can learn how to use warnings and scientific information while contributing their own observations and experience.

Regional Knowledge Anchors in the Middle. Universities and research centers can connect science with communities and local governments, provide technical expertise, and help preserve what is learned over time.

A Learning Network. These different parts must stay connected so that what is observed and learned in one place reaches others, is remembered, and changes what we do before the next disaster.

Learning, in other words, has to travel across places and survive over time.

Looking back across these three essays, I now see three overlapping geographies.

The Geography of Development asks where and how we develop across an interconnected landscape.

The Geography of Risk asks where development choices, together with natural and climate hazards, cause exposure to accumulate.

And the Geography of Learning asks whether what we experience travels across places, survives over time, and changes what we do next.

Ten Years On

I did not think about the problem quite this way when we traveled through Mustang in 2014.

We were thinking about environmental monitoring, universities, communities and the possibilities of learning across a remarkable landscape. Those ideas still matter.

But after a decade of watching Nepal experience repeated shocks—and after thinking more carefully about how information, knowledge and know-how actually accumulate—I find myself returning to that landscape with a somewhat different question.

Ten years from now, Nepal will almost certainly know more about its mountains than it does today. Satellites will improve. Models will become more sophisticated. Artificial intelligence will reveal patterns we cannot easily see today.

But knowing more will not be enough.

The real test is whether a municipal engineer, a teacher, a scientist, a local government and a river-valley community can do things together ten years from now that they cannot reliably do today—and whether what they learn survives, spreads and improves.

Perhaps the larger possibility was never simply to turn the Himalayan landscape into a laboratory.

Perhaps it is to turn it into a Geography of Learning.

Dr. Alok K. Bohara, Emeritus Professor of Economics at the University of New Mexico, writes as an independent observer of Nepal’s democratic evolution through the lens of complexity and emergence science. His systems-policy essays on Nepal’s socio-economic and political landscape appear on Nepal Unplugged.