By: Kabin Maharjan, Eliza Shrestha, Kenisha Shrestha | 1 April 2026
There is now well established climate science which establish climate change causing impacts, losses and damages. But how do we downscale these regional assessments and make a strong case for every event of loss and damage as a case of climate finance? Especially in areas where data are thin but losses are heavy?
The question sounds simple, yet in a data-poor mountain basin, it quickly becomes complicated. Conventional attribution science relies on long-term records and statistical confidence. This approach is biased against countries poor in data infrastructure. What happens when those records are fragmented or absent—exactly in places where people have experienced devastating, irreversible losses?
The Melamchi-Helambu Flood
After the 2021 Melamchi–Helambu flood in a central Nepal catchment, houses of communities were destroyed and livelihoods lost. But this flood was projected as a natural disaster and limited articulation of climate loss.
Of course, a perfect science of attribution is neither desirable nor feasible, but the missing link of climate here is still another form of injustice, excluding already vulnerable regions from evidence, finance, and restorative support.
Why attribution matters, especially in data-poor regions?
Climate attribution has rapidly moved from a scientific frontier to a policy tool. Estimates of climate influence now inform media narratives, policy briefs, and decisions around Loss and Damage (L&D) finance. In the absence of clear attribution narratives, losses and damages risk becoming invisible, not due to their limited severity, but because the data systems needed to legitimise them are often inadequate.
In Melamchi, hydrometeorological records are short, stations poorly capture basin-scale variability, and climate signals interact with steep terrain, landslides, sediment cascades, earthquake legacies, and development choices. These constraints make conventional probabilistic attribution difficult.
For over a year we are working to integrate available data and community stories to establish climate rationale of the Melamchi-Helambu flood (singling out other human and geophysical factors).
For us, this work is not just a technical study, but a justice-oriented effort to ensure that communities living beyond adaptation limits are not excluded from global evidence simply because they live in data-poor regions.
Framing attribution in a cascading flood
Once we committed to doing attribution, a harder question followed: what exactly should we try to attribute?; Do we focus on long-term rainfall/temperature trends, extreme downpour intensity, peak discharge, geomorphic response, or the losses households experienced?
Each framing requires different data, methods, and assumptions, producing different uncertainties. In data-poor settings, these choices determine which stories become scientific evidence and which remain invisible, turning technical framing into political and ethical decisions.
In Melamchi, where flooding unfolded as a cascade of interacting processes, narrowly attributing a single hazard risked oversimplifying reality. Instead, we took a broad and pragmatic approach that remain meaningful even under uncertainty, and relevant to decisions that matter for recovery, finance, and future risk reduction.
Relearning what counts as evidence
Working in Melamchi has forced us to rethink what constitutes as credible evidence when instruments alone cannot carry the full weight of explanation. Rainfall gauges and discharge records provide important fragments, but they do not capture how communities observed changing snow–rain patterns, sediment surges, or shifts in river behaviour over decades.
Rather than treating community knowledge as anecdotal, our work, together with our partner Prakriti Resources Centre (PRC), has approached long-term lived experience as a place-based empirical record, especially valuable where monitoring systems were not adequate. The challenge now is not whether to include these forms of knowledge, but how to integrate them rigorously. This has pushed us to see attribution here as much about epistemic justice as physical causality: whose observations count when evidence is scarce, and which kinds of knowledge are allowed to shape scientific narratives?
For us, rigour means widening the evidence base without lowering standards. By weaving partial climate and hydrological datasets, process reasoning, satellite products, social science methods, and community memory, we are trying to construct an integrated storyline, grounded in multiple forms of evidence.
Beyond single-cause explanations in Himalayan floods
A common misconception is that Himalayan disasters can be explained by a single trigger, mostly rainfall, for instance. Our Melamchi case reaffirmed that they are multi-hazard cascades, where rain interacts with landslides, sediment, channel shifts, and other processes. In such systems, isolating one cause oversimplifies reality and where attribution becomes less about naming a single driver and more about tracing plausible chains of interacting processes.
Disentangling these processes requires an interdisciplinary conversation. Our analysis now brings together glaciologists, hydrologists, climate scientists, and social scientists, each with their own language, analytical frameworks, priorities, and tolerance for uncertainty. At times, we talk past each other, not because anyone is wrong, but because we are optimising for different ways of knowing and application: statistical confidence, causal plausibility, lived legitimacy, or policy relevance.
We are learning to treat this friction not as a failure, but as evidence that the problem is complex enough to challenge everyone’s assumptions. Integrating physical data with community narratives and policy meaning is not a simple add-on; it reshapes the analytical process itself. It is intellectually demanding, but increasingly, it is also where the most productive learning is happening.
The justice implications embedded in attribution
Perhaps the hardest reflection sits at the intersection of attribution and climate justice. Attribution science is increasingly tied to decisions about compensation, accountability, and L&D finance. In principle, it can strengthen claims for support. In practice, it can create a new burden where communities may feel they must scientifically prove climate causality to deserve assistance.
In data-poor mountain regions like ours, uncertainty and limited attribution capacity can be used, sometimes intentionally, to delay action or deflect responsibility. When high evidentiary thresholds become prerequisites for funding, those facing the greatest losses risk exclusion simply because they lack long-term monitoring systems that they never had the power to build. This is not only a methodological challenge but equally a justice dilemma. We see that attribution science matters, but so does how it is interpreted, applied, and embedded within broader climate finance and policy frameworks.
From Melamchi to other data-poor regions: what we are trying to build
We do not yet have a final attribution statement, and we may never produce one that fits neatly within conventional probabilistic language. Yet the insights emerging from the process about method, uncertainty, ethics, and the politics embedded in climate knowledge are no less valuable, especially if they help make the realities of communities and data-poor countries more visible in science and policy. The Melamchi case has shown us that attribution science must evolve, where it must become more contextualised, more innovative, and more attentive to the constraints and complexities of data-poor regions.
The value of our work lies in developing a replicable methodological approach for data-poor regions, where conventional attribution frameworks systematically fall short.
Our aim is to advance an integrated storyline approach that others can adapt across mountain basins and other data-constrained contexts—combining diverse evidence sources, being explicit about uncertainty, and centring questions that matter for justice and finance.
Attribution science will increasingly shape who is seen, who is compensated, and whose losses count. If it is to serve climate justice, it must evolve—becoming more contextual, more inclusive, and more attentive to places where losses are heavy but data are thin.
Acknowledgement
The authors would like to sincerely acknowledge Dr. Hemant Ojha for his continuous support, guidance, and constructive feedback, which have been invaluable in shaping the thinking and reflections presented in this blog. The authors also acknowledge Prakriti Resources Centre (PRC) for their close collaboration and contributions to the field‑based insights and reflections informing this blog.
Disclaimer: The views and opinions expressed in this blog are solely those of the author and do not reflect the positions or views of the organisation with which the author is affiliated or those of the publishers. This blog is intended for general educational purposes only. The author and publishers do not assume, and hereby disclaim, any liability to any party for any loss, damage, or disruption caused by any errors, omissions, or any other cause associated with this blog.
