Anticipating and addressing epidemics - the potential of open data initiatives
Forecast-based Action and the use of OpenStreetMap data
A huge focus in the humanitarian world over the past months has been on epidemics, in particular the COVID-19 outbreak. As infectious disease modelers and epidemiologists work on forecasting disease outbreaks, not dissimilar to how hydrometeorological hazards are modelled and then used in Forecast-based Action (FbA), we want to explore how these two worlds can be brought together to strengthen early action for epidemics.
For a FbA system to flourish, key elements should be in place, namely: triggers, actions and finances. An essential part of the decision-making process to decide when, where and why to trigger early action is the use of impact-based forecasting to create so-called early action intervention and activation maps. In order to conduct this forecast and risk analysis, comprehensive and good quality geospatial data is required which covers aspects of vulnerability, exposure and hazard. This is where OpenStreetMap (OSM) comes in, in particular Humanitarian OpenStreetMap Team (HOT) and the Missing Maps project.
OSM was founded in 2004 in the UK with the goal to create a free and open source map of the world. Initially OSM activities were centred on developed countries, but over the years many activities started to focus on developing nations.
The 2010 earthquake in Haiti was the beginning of the journey of humanitarian mapping in OSM. The Humanitarian OpenStreetMap Team (HOT) was established that year to promote and support the work of thousands of volunteers and dozens of local groups around the world that contribute to OSM, with an emphasis on humanitarian mapping to aid disaster relief, preparedness and economic development. In its role, HOT organizes crisis response mapping and acts as a bridge between the OSM community and traditional humanitarian actors like Médecines Sans Frontièrs (MSF), the Red Cross Red Crescent Movement, UN Agencies and NGOs, as well as local groups and governments.
In 2014, the Missing Maps project was founded by the American Red Cross, British Red Cross, HOT and MSF. It now has a total of 19 members, including the IFRC and the German Red Cross. The aim of Missing Maps is to map the most vulnerable places in the world. Missing Maps conducts pre-emptive mapping in OSM, mapping areas before a disaster occurs.
Missing Maps is supporting FbA, making OSM more and more essential for the process of deciding when, where and why to trigger early action. OSM exposure data such as houses, schools, hospitals, roads, water points, bridges, etc. are very important to understanding what is likely to be impacted by a hazard. Of course, this depends on the vulnerability of these buildings, infrastructure and the people who interact with them. Many parts of the world have very limited data of sufficient quality on exposure and vulnerability. To address this gap, OSM offers open access data that can be used in combination with weather and climate forecasts (and other data) to move towards the future of forecasting: “Impact-based forecasting”. This is one of the key elements of a FbA system to trigger early action to protect the lives and livelihoods of the most vulnerable people in the world.
Currently, FbA projects in Peru and Indonesia are using OSM data, flood forecasts and other data to trigger early action before potential dangerous floods occur. In both countries, Missing Maps projects have been instrumental in expanding the map to areas that are at high risk of floods but that were previously unmapped. This strengthens our capacity to act early. There is still more work to be done to expand the map to cover all the potential communities that could be affected by growing risk, particularly in the face of a changing climate, armed conflicts and increased vulnerability due to COVID-19.
FbA and Epidemics
Missing Maps members have been laying the groundwork for a potential use of FbA for epidemics in their involvement in COVID-19 responses. Some of their activities are presented below.
Since March, the HOT community and their partners have worked across three continents to provide data needed for COVID-19 responses, mapping over 227 locations, engaging over 10,000 volunteers to date and providing eight Microgrants to local communities to map in support of COVID-19 responses in their own countries.
In responding to COVID-19, HOT has identified three key areas of support:
- Help government agencies and responders with data so they can provide essential services to all communities, including those in rural and under-mapped areas
- Identify populations most at risk by identifying and addressing data gaps on risk factors such as age, pre-existing conditions and access to public health services and public gathering and resource distribution centers like markets that households may depend on for food and livelihoods
- Increase map creation in at-risk locations by supporting local mapping communities in creating maps that local responders can use
The types of features being mapped vary based on the local needs, but can include building footprints, roads, clinics, handwashing stations, markets and other gathering places, place names and administrative boundaries. Together, these items support the priorities above by enabling governments and communities to reach previously unmapped communities and households, gauge the risk of both the pandemic and the side-effects of lockdowns like loss of livelihoods, plan public health interventions and help individuals to find essential services.
In Botswana, HOT has been mapping at the request of the Botswana Institute of Geoinformatics to support the national COVID-19 response. While the mapping of the initial priority region has been completed, efforts are now expanding to map the entire country in OSM.
In Mali, HOT is supporting the COVID-19 coalition, which includes members of the local OpenStreetMap community and the Mali Red Cross, so they can better plan and respond to coronavirus impacts.
In Peru, HOT has mapped the entire Cusco region, home to 1.2 million people and with extremely varied terrain; from Altiplano communities living at 5,000m altitude to Amazonian jungle. Creating better maps has allowed government workers to begin using the data to assess healthcare needs. This is particularly vital for maternal healthcare services, which have been largely unavailable over the past few months due to quarantine.
The Philippines OSM community has mapped Quezon City, one of the most densely populated places in Manila and the hardest hit by COVID-19 at the time the project started. This has been in partnership with the Pacific Disaster Center as part of their project to map critical lifeline infrastructure across the country.
In addition, local OSM communities and partner organizations have been setting up mapping projects around the world, with 20 currently underway on the HOT Tasking Manager, mapping their own locations and interests regarding the coronavirus
Awareness of local health capacity is of the utmost importance when planning and conducting humanitarian projects and in disaster response contexts in general. This again became very apparent in the current global COVID-19 response.
The Heidelberg Institute for Geoinformation Technology (HeiGIT gGmbH), also a Missing Maps member, was launched to improve knowledge and technology transfer from fundamental research in geoinformatics to practical applications with a specific focus on user-generated geodata, e.g. OSM, and environmental and humanitarian contexts. In this vein, HeiGIT develops related applications, workflows and approaches.
To enable awareness of local health capacities and building on the data created through HOT efforts, in the scope of the COVID-19 response the HeiGIT team developed an extension of their OSM History Explorer (ohsome hex) which allows analyses of OSM completeness in mapping health facilities in Sub-Saharan Africa. This extension facilitates quick and simple assessment of health facilities in areas of interest and to respectively raise awareness of vulnerable areas. Additionally, the team conducted several related research studies and developed a “Map of Hope” that provides a comprehensive overview of COVID-19 related clinical trials, using OSM as a base layer for localization.
Both HOT and HeiGIT have demonstrated the power of leveraging OSM in responding to the COVID-19 outbreak. The newly created OSM data has played a crucial role in better planning and responding to COVID-19 impacts, and invaluable lessons have been learned from this that could help in building the first steps of early action for future epidemics.
For one, location data is not the only kind of data needed for a useful map. Data users working to address pandemics need to know not only where places like markets and water points were located, but also details such as user capacity and hours of operation in order to fully gauge risks.
In fact, in many places where this kind of detailed data about the availability, use, and risk of services, resources, and public spaces is needed, the general base map of roads and buildings is already largely complete. Due to this, the greatest need for volunteers is often in local mapping rather than remote, digital mapping. HOT has found it invaluable to provide support to existing local mapping communities in purchasing equipment, accessing training and the like so they could address local mapping needs more effectively.
There is also a need to develop long-term partnerships in anticipation of a need to generating open map data. There are countries that are currently heavily impacted by COVID-19 and unmapped in OpenStreetMap, but where there have not been requests for assistance in mapping. There is an opportunity need to develop relationships with government at all levels, civil society, and volunteer mappers, in order to be well-positioned to providing assistance with mapping for pandemic response in the future.
OSM offers an immense opportunity to manage risks in a new way by considering the nature of compound and cascading hazards and where climate, epidemic, conflict and other hazards collide. The better the map, the better our capacity to identify who and what could be impacted. This allows the development of tailored solutions in the short, medium and long term. Both digital and in-person Red Cross Red Crescent volunteers have been playing an enormous role in enhancing OSM data. By increasing our digital volunteer capacity to map remotely in mapathons and by ensuring that local mapping and validation is done, we will be able to increase and enhance the amount of crucial information that is available not only for anticipatory action but in general to adapt and reduce risks at all times.
Volunteers have been the foundation of the RCRC movement since it’s beginning and are a central part of all activities. We, therefore, encourage the RCRC movement to invest in volunteers’ capacity to map all over the world and across all different types of programmes, from health, to shelter, to DRR and livelihoods. By doing so the benefit will be multiplied to make the difference that is really needed and fulfil our 2018 goal of leaving no one behind.