Developing an impact-based forecasting model within Nepal’s national disaster information management system: the BIPAD portal
Nepal faces multiple disasters which arise predominantly from climate-related hazards during the summer monsoon period. Heavy rainfall often brings widespread flooding, while landslides result in huge losses of life and property across the country.
There has been significant progress in flood forecasting and numerical weather prediction in recent years, but the communication and dissemination of these – to stakeholders and the public – remains vague, and there is little application of these forecasts for early action. Conventional approaches to hazard warnings are becoming unfit for purpose, unable to cater to the increasing demands for early intervention across Nepal’s growing anticipation community.
In this context, the UK Met Office and Red Cross Red Crescent Climate Centre, under the Asia Regional Resilience to Changing Climate programme, have been working with the government of Nepal, especially the National Disaster Risk Reduction and Management Authority and the Department of Hydrology and Meteorology to pilot impact-based forecasting in Nepal.
In March 2021, the Met Office and the Climate Centre jointly delivered stakeholder training on ‘Impact-based forecasting for early action’ in Nepal. Since then, there has been significant progress. The government is now exploring different ways to integrate impact-based forecasting in their warning services and decision-support system for monsoon-related hazards. One recent development is the development of an impact-based forecasting module within the BIPAD portal.
The BIPAD portal
The BIPAD portal – Building Information Platform Against Disaster – was initiated by the Ministry of Home Affairs and is owned by the National Disaster Risk Reduction and Management Authority. BIPAD is the national disaster information management system in Nepal – but with separate platforms for national, provincial and municipal governments.
The portal currently has six modules – dashboard, incident, damage and loss, real-time, profile, risk information – which aim to support decision-making across all phases of disaster management. Working with the Youth Innovation Lab, the technical team working on the BIPAD portal, the Red Cross Red Crescent Climate Centre and the Nepal government have now developed a further module focused on impact-based forecasting within the portal.
Forecasting the impacts of floods
The impact-based forecasting module of the BIPAD portal currently focuses on riverine floods and is being piloted at two river stations in West Nepal: Chisapani station on the Karnali River, and Chepang station on the Babai River, which cover the flood plains of Bardiya and Kailali districts respectively (Figure 1). The portal integrates hydrological forecast data from the Global Flood Awareness System (GLoFAS) for these two stations, and interlinks with METEOR flood inundation maps at different return periods, to calculate and visualize flood impacts.
The portal combines this flood hazard data with risk data available at different spatial scales (e.g., on vulnerability, coping capacity, exposure). These are used to visualize quantifiable potential impacts of forecast flood events in near real-time. All data and information are presented interactively, and as simple visualizations, to help end-users understand the estimated impacts – and then act in anticipation of these impacts.
Although the portal currently integrates global flood forecast with lead times of up to 10 days, it is flexible enough to integrate local flood forecasts from the Department of Hydrology and Meteorology. This enables BIPAD to quantify the potential level of impact associated with flood warnings with different (i.e., shorter) lead times.
Visualizing and quantifying the expected impact of floods
The impact-based forecasting module quantifies and visualizes three main elements in a dashboard. The first is potentially exposed elements in flood-prone areas, including buildings, roads and land use. The data is extracted from OpenStreetMap for different administrative levels (i.e., district and municipality), which is overlayed onto flood extent maps for the forecast event to quantify the expected impact (Figure 2).
The second element is at-risk households. This is quantified for five flood-exposed municipalities (Tikapur, Janaki, Gulariya, Bara Bardiya, Thakurbaba) in the Kailali and Bardiya districts, using detailed household-level risk data collected by the Nepal Red Cross Society under the ‘Forecast-based action and shock responsive social protection’ project. The portal quantifies a household-level risk score using three dimensions: (1) hazard and exposure; (2) vulnerability; and (3) lack of coping capacity. Following the INFORM Risk Index methodology, several indicators around these three dimensions are evaluated and weighted to quantify the underlying risk score, which are then categorized into five classes: very low (0-1.9); low (2-3.4); medium (3.5-4.9); high (5-6.4); and very high (6.5-10). These are visualized using different colours (Figure 3). The system also allows users to visualize the household score for each dimension separately, along with the types of parameter used to determine risk (Figure 4).
Figure 3. Representation of household-level risk scores; each dot representing a single household, which is given a colour based on its score. Source: BIPAD portal
Figure 4. Visualization of types of indicator used for determining risk. Clicking an individual indicator in the overview pane updates the household data shown in the map. Source: BIPAD portal.
Figure 3. Representation of household-level risk scores; each dot representing a single household, which is given a colour based on its score. Source: BIPAD portal
Figure 4. Visualization of types of indicator used for determining risk. Clicking an individual indicator in the overview pane updates the household data shown in the map. Source: BIPAD portal.
The third element that the impact-based forecasting module of BIPAD portal quantifies and visualizes is the potential impacts of forecast floods at the household level. The flood magnitude is combined with risk scores to determine the potential severity of impact for each household (Figure 5). Red indicates households that could report a very high impact (if the forecast flood materializes), while green represents households that will experience a potentially low impact.
Disaster management and humanitarian agencies – including the National Disaster Risk Reduction and Management Authority, the Nepal Red Cross Society and local authorities – can use this risk and impact information to identify households expected to be severely affected, and direct efforts towards them before the impacts of a flood materialize. In this way, the BIPAD portal can potentially provide the information needed to trigger a nationwide anticipatory action approach to river floods in Nepal.
Future work
The BIPAD portal is the first to quantify and visualize the potential impacts of floods by overlaying forecast data in near real-time with exposure and vulnerability data in Nepal. However, the impact-based forecasting module is still a prototype, with limited practical functionality. It currently displays only the very basic requirements of an impact-based forecasting model, with little user interaction to date. Through continuous user testing and piloting at the local level, the impact-based forecasting module is expected to be fully operational for the 2022 monsoon season.
There remains a need to integrate more reliable and locally available flood forecast data into the portal, along with flood extent maps. Discussions are ongoing about the integration of local forecast produced by the flood forecasting division of Department of Hydrology and Meteorology, which are often more reliable and accurate for shorter lead times.
More broadly, scaling up the system will depend on the availability of risk databases with socio-economic vulnerability information at different levels. These data are often limited and in most cases only available on a project basis, meaning they remain within organizations and are rarely available for use. The National Disaster Risk Reduction and Management Authority is coordinating with several agencies, both government and non-government, to collate the available risk information in the BIPAD portal, which can later be integrated into the impact-based forecasting module to cover more regions.
The National Disaster Risk Reduction and Management Authority also aims to scale up this portal to cater for multiple hazards and different types of risk information. Ultimately, BIPAD will evolve into a decision-support tool for wider risk-reduction efforts and anticipatory action across Nepal. This initial prototype is a stepping stone towards achieving this much larger goal: of risk-informed planning and action in disaster management and humanitarian operations.
This blog was written by Madhab Uprety, Climate Centre, Rajendra Sharma, National Disaster Risk Reduction and Management Authority, and Reena Bajracharya, Youth Innovation Lab.