The new study builds directly on a substantial body of earlier work by the same research group. In previous publications, the team demonstrated the performance of statistical models including the Frequency Ratio, Information Value, and Shannon Entropy methods for producing landslide susceptibility maps of Bhutan. These models quantify how strongly factors such as slope, lithology, land cover, proximity to roads and rivers, and rainfall correlate with observed landslide occurrences, and they convert those correlations into spatial predictions of where future slides are most likely. The researchers then extended the susceptibility work into a full landslide risk index, adopting the widely used conceptual framework in which risk is the product of three dimensions: hazard, meaning the likelihood and intensity of landslide occurrence; exposure, meaning the people, buildings, infrastructure, and assets located in hazard-prone areas; and vulnerability, meaning the social and physical susceptibility of those exposed elements to harm. That national-level risk index was computed using an indicator-based approach for every Gewog in the country, and it is precisely this risk framework that the new web platform now makes operational and interactive.

Technically, the platform is constructed from a modern, service-oriented geospatial architecture. ArcGIS Enterprise serves as the backbone, hosting the spatial layers that depict hazard zones, exposure datasets, and vulnerability indicators as map services that can be streamed to any web browser. Tabular data, including the indicator values and attribute information associated with each Gewog, are stored in a PostgreSQL relational database, which provides a robust and open-source backend for querying and managing structured records. The front end is developed with standard web technologies: HTML provides the document structure, CSS controls the visual presentation and layout, and JavaScript drives the interactivity, with the ArcGIS JavaScript API supplying the mapping widgets, layer controls, and client-side rendering that allow users to pan, zoom, toggle layers, and interrogate features directly within a browser. This combination means that no specialized GIS software or expert training is required to use the system; anyone with an internet connection and a standard web browser can access the country’s landslide risk information.

What distinguishes the platform from a static risk map is its emphasis on contextual information and analysis. Users can explore each of the three risk dimensions separately or view them integrated into a composite risk picture at the Gewog level. Beyond the headline risk classification, the system offers contextual analysis of a range of social, physical and infrastructural, and environmental indicators, allowing a decision maker to understand not just that a particular Gewog is at high risk, but why. A planner might discover, for example, that a Gewog’s elevated risk score stems from a combination of highly susceptible slopes and a dense network of roads and settlements cutting across them, or that a moderate hazard level is amplified by a vulnerable population with limited access to services. This layered, interrogable presentation of evidence is precisely what indicator-based risk frameworks are intended to support, and the platform operationalizes that intent for a national audience.

The collaborative dimension of the system is as important as its analytical content. The authors argue that a collaborative effort and an open information platform are crucial for enhancing long-term landslide risk reduction, particularly at the micro-level, and for effective land-use planning. By integrating efforts from multiple government departments with relevant expertise, rather than relying on isolated, locally stored datasets, the platform ensures that landslide-related information remains consistent and up to date. In many countries, disaster risk information is fragmented across agencies that maintain their own copies of overlapping datasets, leading to inconsistencies that undermine planning decisions. A centralized web service architecture addresses this problem directly: when a contributing department updates a dataset, the revised information flows through the same hosted services to every user of the platform, eliminating version conflicts and redundant data maintenance.

The research team behind the system reflects an institutional cross-section of Bhutanese geospatial and engineering expertise. Lead author Indra Bahadur Chhetri of the Department of Civil Engineering and Surveying at Jigme Namgyel Engineering College, Royal University of Bhutan, conceived the core idea, developed the methodology framework, analyzed the spatial content, and led application development and manuscript preparation. Co-authors Tshering Dorji Sherpa of Druk Green Power Corporation Limited, Menuka Rai of the National Land Commission Secretariat, and Sonam Jamtsho and Yonten Jamtsho of Rigsar-Vajra JV Private Limited contributed to data collection and the coding of the application. The spread of affiliations, spanning academia, a state power utility, the national land administration, and a private surveying firm, mirrors the multi-stakeholder philosophy embedded in the platform’s design, in which different organizations supply and consume risk information through a shared infrastructure.

Bhutan’s landslide problem provides the urgent context for this work. The country’s terrain is among the most landslide-prone in the Himalayan region, and the hazard intersects directly with national development priorities, since roads, hydropower infrastructure, and growing settlements must all be planned and maintained in steep, geologically young, and tectonically active landscapes. Previous research by other groups has documented the extent of the challenge, including spatial landslide risk assessments at Phuentsholing, susceptibility mapping along the Asian highway corridor, landslide identification using the index of entropy technique, and UAV-based localization of specific failure sites such as the Rinchending Goenpa landslide. Studies of roadblock events caused by geohazards have highlighted the disruption that slope failures inflict on transportation and the economy. Against this backdrop, a national, Gewog-level, web-accessible risk platform fills a clear operational gap between academic risk modelling and the day-to-day decisions of land-use planners and emergency responders.

The Bhutanese platform also enters a well-established international conversation about the role of web-based GIS in disaster risk management. Prior systems have demonstrated the value of the approach in other hazard contexts: dynamic web-GIS landslide early warning systems have been developed for the Chittagong metropolitan area in Bangladesh; web-based GIS platforms have been used to manage and assess landslide data in Peace River, Canada; collaborative web mapping applications built on REST API services and open data have been deployed for landslides and floods in Italy; and integrated two- and three-dimensional WebGIS platforms have been proposed for landslide hazard management more broadly. Systematic reviews of Web-GIS for natural hazard management confirm a growing consensus that web services are the most scalable way to deliver geospatial risk information to non-specialist users. The Bhutanese system contributes a national-scale, indicator-based, three-dimensional risk framework to this landscape, tailored to the administrative structure and institutional realities of a small mountain kingdom.

The design choices embedded in the system also align with international policy frameworks. The United Nations Office for Disaster Risk Reduction’s Sendai Framework for Disaster Risk Reduction 2015–2030 calls for understanding disaster risk as the first priority for action, and emphasizes the role of accessible, disaggregated, and up-to-date risk information in supporting evidence-based policy. Multi-stakeholder platforms have long been recognized in the disaster risk reduction literature as mechanisms for adaptive governance and resilience. By providing a shared informational foundation on which multiple organizations can coordinate, the Bhutanese WebGIS platform functions as a practical instrument of such governance, translating the abstract goals of risk-informed development into a concrete tool that can be consulted before a road is realigned, a settlement expanded, or a monsoon-season response planned.

The authors report no competing interests and note that the research received no external funding. The data supporting the findings can be provided by the corresponding author upon reasonable request. While the platform as described is a two-dimensional system, the broader trajectory of WebGIS development, including three-dimensional visualizations of seismic and landslide risk demonstrated elsewhere, suggests possible avenues for future extension. For now, the significance of the work lies in its demonstration that a country with limited resources can assemble a modern, standards-based geospatial decision support system from open web technologies, a relational database, and hosted GIS services, and that doing so can change the institutional economics of disaster risk information. Instead of each agency maintaining its own brittle collection of files, the risk knowledge becomes a living, shared service. In a region where climate change is expected to intensify the rainfall triggering mechanisms behind landslides, and where every monsoon season brings new slope failures, the value of such a shared, current, and analytically rich picture of risk is difficult to overstate. The platform represents a step toward making landslide risk management in Bhutan proactive, collaborative, and continuously informed by the best available evidence, offering a model that other hazard-prone nations in the Himalayas and beyond may find instructive.

Landslides represent a perennial threat in many mountainous and steeply sloped regions, frequently exacerbated by antecedent weather conditions and seismic activities. While previous models mainly focused on single factors such as rainfall intensity or seismic tremors independently, the new approach developed by Zeng, Zhang, Xiao, and their collaborators delves into the coupling effects—that is, how rainfall and earthquakes jointly influence slope stability. Such synergy between external forces has long been suspected to elevate landslide risks but remained challenging to quantify until now.

The essence of the breakthrough lies in the refined assessment framework that integrates hydrological dynamics with seismic shaking parameters into a cohesive predictive algorithm. By incorporating real-time and historical datasets related to precipitation patterns and earthquake magnitudes, the model calculates a susceptibility index that is remarkably sensitive to the fluctuations induced by combined triggering mechanisms. This dual-factor methodology represents a paradigm shift from conventional monovariate risk models to a more holistic multifactorial risk assessment tool.

From a technical standpoint, the model leverages advanced statistical techniques and machine learning algorithms that interpret nonlinear interactions between rainfall-induced pore water pressure elevations and earthquake-generated ground accelerations. The researchers harnessed geospatial information systems (GIS) to pattern these hazard interactions over diverse geomorphological landscapes, enabling high-resolution susceptibility mapping. This precision opens new avenues for proactive hazard identification and emergency response prioritization.

Critical to this approach is the incorporation of soil mechanics principles, notably the reduction in shear strength caused by rainfall infiltration, which primes slopes to fail when subsequently jostled by seismic waves. The model quantifies this weakening effect through parameters like hydraulic conductivity and soil cohesion, meshed with earthquake shaking intensity measures such as peak ground acceleration (PGA) and spectral acceleration values. This scientifically rigorous coupling framework reflects a deeper mechanistic understanding than prior empirical models.

The research team validated their model against well-documented landslide events in regions prone to both intense tropical rainfall and frequent seismic activity. In these validation exercises, their integrated model significantly outperformed traditional models by accurately predicting landslide occurrence with higher recall and precision rates. Such validation underscores the model’s robustness and practical applicability in real-world hazard mitigation programs.

One of the striking implications of this research is its potential utility in early warning systems. By continuously monitoring rainfall accumulation and seismic activity indicators, authorities could deploy this model in near real-time to forecast landslide susceptibility spikes, allowing timely evacuation orders and infrastructure safeguarding measures. This capability could revolutionize disaster management, minimizing loss of life and economic damage in vulnerable communities.

This multi-hazard modeling also invites a reexamination of current land use policies, especially in regions undergoing rapid urban expansion into hilly terrains. The refined susceptibility maps can guide planners to avoid highly risky zones or implement engineering controls such as slope reinforcement and drainage improvements in susceptible areas. Consequently, infrastructure resilience could be enhanced in disaster-prone regions around the globe.

Moreover, the implications extend beyond immediate hazard prediction. The framework proposed by Zeng and colleagues opens new possibilities for climate change impact studies, given projections for increasing rainfall variability and seismic risks induced by anthropogenic activities. Understanding the coupled effect of these natural forces equips scientists and policymakers with a predictive lens to anticipate evolving geohazard landscapes in a warming world.

From a scientific methodology viewpoint, the integration of real-time sensor data into the dynamic version of the model promises significant progress. As sensor networks measuring both hydrological and seismic parameters continue to expand, this could fuel continuous model updates enhancing prediction accuracy. The adaptive learning components embedded in the model are primed for such data streams, marking a future direction brimming with potential.

Despite the breakthrough, the authors acknowledge that challenges remain. Data scarcity and variability in some mountainous regions can limit the immediate applicability of the model, calling for enhanced monitoring infrastructure. Furthermore, calibrating the model to account for local geological heterogeneities, vegetation cover effects, and anthropogenic modifications requires ongoing research efforts.

The contribution of this refined coupled rainfall-earthquake landslide susceptibility model stands as a compelling example of how interdisciplinary research—merging geotechnical engineering, hydrology, seismology, and data science—can yield transformative tools for environmental risk management. It exemplifies the kind of integrative thinking needed to confront the multifaceted nature of natural disasters in the 21st century.

Looking ahead, the authors suggest expanding the framework to incorporate other potential landslide triggers, such as snowmelt and human activities like mining and deforestation. Furthermore, coupling this approach with socioeconomic vulnerability assessments could lead to comprehensive disaster risk reduction strategies that not only identify hazards but also focus on human resilience.

In summary, this innovative research marks a significant leap toward sophisticated, multi-dimensional natural hazard modeling. By capturing the nuanced interactions between rainfall and seismic forces that precipitate landslides, it equips communities and governments with sharper tools to foresee, prepare for, and ultimately mitigate the impacts of these devastating events. As climate change and urban pressures continue to reshape vulnerable landscapes, advancements like this underscore the importance of science-led strategies to safeguard lives and livelihoods.

Subject of Research: Landslide susceptibility modeling considering the coupling effects of rainfall and earthquakes.

Article Title: A refined assessment model for landslide susceptibility under rainfall-earthquake coupling effects.

Article References: Zeng, Y., Zhang, Y., Xiao, S. et al. A refined assessment model for landslide susceptibility under rainfall-earthquake coupling effects. Environ Earth Sci 84, 662 (2025). https://doi.org/10.1007/s12665-025-12552-2

Image Credits: AI Generated

DOI: https://doi.org/10.1007/s12665-025-12552-2