News — 27 March, 2023

Changing Data Needs Through the Disaster Management Lifecycle

We’re now about a month and a half after the initial 7.7 magnitude earthquake hit Turkey and Syria on Monday, 6 February. Even though it is an ongoing emergency with vast humanitarian needs, we are now moving into a phase of continuing support to the affected communities and the start of rebuilding and reconstruction efforts where it is feasible to do. As the phase of disaster response changes, so do the data needs and data and quality requirements.

Data Availability and Uses: OSM Data Across the Disaster Management Phases

During each phase of the ‘disaster management cycle,’ there are different data needs, uses, and requirements. Let’s first take a brief look at the different phases.

Response activities immediately following a disaster event to help affected people include life-saving search and rescue and meeting basic humanitarian needs: water, food, clothing, medical aid, shelter, etc.

Initial data activities involve assessing OSM data availability and gaps, procuring as much pre- and post-event open-source imagery as possible (satellite and UAV), and supporting assessments of damage and humanitarian and logistical needs.

Response-related Datasets:

Dataset	Data needs and use	Source
Buildings	Population estimates and #/% affected.	OpenStreetMap
Roads	Routing and logistics.	OpenStreetMap
Place names	Routing and navigation.	OpenStreetMap
Event data	Area of impact & severity to estimate the affected population.	GDACS
Context-specific infrastructure data	Examples include health facilities, evacuation sites, airports, and helicopter landing sites. What’s relevant and needed depends on the type of disaster event and the context.	Mix of OSM, HDX data grids, third party datasets
Programmatic and needs data	Humanitarian needs assessments, program activities, etc.	Third-party datasets

Recovery activities focus on returning the situation to normal (and ideally better and more resilient) conditions and circumstances for affected people. The focus is still meeting the basic needs of the affected population, but it is also shifting to more thorough safety and damage assessments, cleanup, and rebuilding and reconstruction.

Recovery-related Datasets:

Dataset	Data needs and use	Source
Buildings	Population estimates and addressing - the ability to analyze with other datasets.	OpenStreetMap
Roads	Routing and logistics for the ongoing delivery of humanitarian aid.	OpenStreetMap
Damage assessments	Combining OSM data with external & third-party datasets to assist in damage assessments.	UNOSAT, Copernicus, etc.
Context-specific infrastructure data	What’s relevant and needed depends on the type of disaster event and the context. This could be for education, healthcare, markets, financial services, or other areas.	Mix of OSM and third-party datasets
Programmatic data	Humanitarian and development needs assessments, program activities, etc.	Third-party datasets

Preparedness activities include planning and preparation to improve local capacity and resilience to cope with future events and emergencies. Often, a separate “mitigation” phase is defined between recovery and preparedness, focusing on prevention and reducing the likelihood of effects and impacts. In this case, when looking at (geospatial) data needs, we’re combining mitigation with preparedness.

Preparedness-related Datasets:

Dataset	Data needs and use	Source
Key infrastructure (context-specific), such as water infrastructure and drainage, public transport, or health and education	Specific interventions in an area to improve resilience and preparedness.	OpenStreetMap plus government or third-party datasets
Natural environment: such as wetlands, floodplains, and mangroves or river systems	Understand a wider ecosystem and the impacts of interventions and investments.
Project specific datasets	Specific interventions in an area to improve resilience and preparedness.
Evacuation sites and coping capacity	Creating evacuation and escalation planning.

You can access OSM data via HOT’s HDX page.

Data Requirements and Quality: Approaching Evolving Data Quality Needs at HOT

As data use differs for each phase of a response, so do the expectations and requirements of the data itself. Often, higher data quality standards are expected of data in ‘stable’ locations and times than in the immediate disaster response phase.

Response Phase

During the immediate Response Phase, data availability is key - and data quality is often good enough as long as it doesn’t hamper the primary use of the data. For example, while annoying, overlapping or non-square buildings don’t actually cause population estimates to be significantly off. However, major roads that don’t connect cause issues for route logistics and navigation.

Comparison of mapping from 10 February (left) to 22 March (right). Light blue denotes areas of mapped buildings in OSM.

Comprehensive data can be scarce and initially hard to get, especially if we’re dealing with an area that hasn’t been mapped well in the past. The pace of mapping on the Tasking Manager will invariably outpace data quality checks and validation.

To complement tasking and data creation, we run a Data Protection Risk Assessment to ensure we consider local risk factors and potential harms of mapping, and consider potential improvements and mitigations. For example, we are careful of issues around the representation and exclusion of rural populations or ethnic groups through our tasking areas and definitions and what types of infrastructure we request (or don’t request) to be mapped.

Recovery Phase

During the Recovery Phase, our focus will shift more toward putting a concerted effort into data cleanup and validation. Also, we may be supporting specific recovery efforts, such as more comprehensive damage assessments, providing ongoing humanitarian assistance to shelters and camps, and planning for reconstruction efforts. This does require data to be higher quality overall and link to or incorporate additional data items, such as detailed addressing.

Preparedness Phase

For the Preparedness Phase, high data quality and fidelity is required - especially when considering the implementation of specific projects for disaster risk reduction, resilience, or anticipatory action. Often, data collection and mapping outputs will go towards providing decision-making data and baseline information on specific interventions and investments. Examples of this kind of project can be found in the Open Cities program (for example, in Accra, Ghana). These projects aim to build the disaster resilience of the cities they take place in and require detailed and high-quality data on specific aspects of infrastructure, such as the exact layout and capacity of drainage and water systems, public transport systems, or health facilities.

Want to contribute to OSM for humanitarian needs? Visit HOT’s Tasking Manager for projects starting at the beginner level, and read HOT’s Tips for Mappers. You can access OSM data via HOT’s HDX page.