When a disaster occurs, governments grapple with many questions: How bad is the damage? What are the priorities for response and recovery? What are the impacts on housing and infrastructure? And where are the impacts greatest?
For The Bahamas, following the devastating impact of Hurricane Dorian, the questions were no different. Hurricane Dorian made landfall in The Bahamas as a Category 5 storm on September 1, battering the northern islands of Grand Bahama and Abaco for almost three days. As of October 28, reports indicated that 67 people lost their lives and 282 people remained missing.
In the days following the hurricane, reports clearly showed devastation but left many questions. The World Bank’s Global Rapid post-disaster Damage Estimation (GRADE) approach, developed by its Disaster-Resilience Analytics and Solutions (D-RAS) team, was able to address many of the critical questions within 13 days using a rapid, desk-based approach.
What is GRADE?
GRADE provides an estimate of physical damages to housing and critical infrastructure using a methodology that considers the:
- Hazard (e.g. Intensity of windspeeds)
- Exposure (E.g. Value and location of buildings and infrastructure)
- Vulnerability (E.g. Extent and cost of damage to exposed assets)
The D-RAS team then combines this information, checks it carefully against observations and historical precedent and applies expert knowledge to produce results within two weeks of the commencement of the analysis.
In the last five years, this methodology has been applied in 15 countries, covering earthquakes, windstorms, floods and volcanic eruptions worldwide.
Figure 1: Previous GRADE analyses conducted by the World Bank’s Disaster-Resilience Analytics & Solutions (D-RAS) team. Red lines indicate storm tracks.
How was this Applied in The Bahamas?
Each GRADE analysis is challenging and taxing, but The Bahamas analysis was particularly so. If we consider an 8-hour day as a typical work-day, the GRADE team clocked over 15 hours per day in the initial phase of GRADE, to produce results within 13 days. The D-RAS team included six experts in hazard mapping, risk exposure, vulnerability modeling and validation, and comparative historical disaster analysis. Challenges experienced include:
- Variation in building types. Buildings ranged from high-value properties and resorts to informal settlements, sometimes side-by-side.
- Complexity of the perils. Damage was due to wind, wind-driven rain, flooding caused by the rain, and storm surge, all acting intensely over several days.
- Lack of access to reliable, up-to-date data. Reports conflicted on the levels of damage, making the cross-checking process particularly difficult: Insured loss estimates from the private sector ranged from USD 1.5bn to over USD 6bn; estimates derived from Artificial Intelligence showed discrepancies with the damage we were seeing; and on-the-ground assessments were ongoing.
However, this event had lots of aerial imagery to draw upon, so our damage estimates were strengthened by assessments conducted through satellite imagery, aerial flyovers, drone and ground-based imagery, along with social media feeds.
What Were Our Main Findings?
The analysis revealed that:
- [The GRADE report provides a detailed breakdown of damages for the residential, non-residential, tourism and infrastructure sectors.]
- The event particularly affected two islands and several cays – the Abacos and Grand Bahama - where around 20% of the Bahamas population lived. Although Grand Bahama has a significantly higher value of buildings and infrastructure, the Abacos suffered proportionally more damage.
- The spatial distribution of damages was highly unevenly distributed, with some areas experiencing significant destruction, and others experiencing minimal impact.
- Good quality housing, built to effective wind codes, often suffered much less damage compared to their less-well-constructed neighbors. Also, informal settlements on the Abacos, called The Mudd and The Pigeon Peas, had far less resiliently constructed buildings, and the settlements were completely devastated by the force of the wind and the ensuing storm surge.
Consideration therefore needs to be given to how to improve construction practices and adherence to building codes, and how zoning policies can be used to reduce losses in future events.
Why is this information useful?
The economic damage estimated by specialized companies related to the re/insurance sector can also be highly varied. Here, it is important for the World Bank to have its own rapid estimate to enable decision-making on the best means to support affected countries.
to inform ongoing efforts to recover from this disaster and build future resilience.
Figure 2: How the GRADE method was applied to Hurricane Dorian. Key components are shown in white, and outputs are shown at bottom.
[This blog received inputs from D-RAS Team members: Antonios Pomonis, James Daniell, Graeme Riddell, Josh Macabuag and Andreas Schaefer.]