Like other coastal cities, Honolulu’s long-term viability depends on how well it can adapt to climate change. By 2100, sea levels are predicted to rise by at least a meter worldwide alongside an increasing number of extreme weather events.
How the one-two punch of sea-level rise and more flooding hazards will affect people and infrastructure cannot be viewed in isolation but must be seen as part of a broader pattern. Sea level rise exacerbates flooding in coastal areas, which can then be compounded by extreme weather such as storm surges, tsunamis, and even fires. Policymakers and planners in Honolulu and elsewhere must understand plausible worst-case scenarios to “prepare for the worst and hope for the best.”
To understand what the worst would look like, we quantified flooding risks in Honolulu due to sea level rise and severe weather events and assessed the risks and vulnerabilities of critical infrastructure. This included threats to the city’s transportation system, for which we used travel demand software to model evacuation from inundated areas. Our approach provides critical information for emergency responders, transportation systems managers, planners, and developers that inform potential policies to protect the city and its residents.
The risks associated with flooding can be estimated by probability or more precisely through modeling historical events. We used both approaches. Over the long term, certain areas are susceptible to flooding even without any particular hazardous event. In April 2018, for instance, the island of Kauai recorded about 50 inches of rain during a 24-hour period, which is currently being certified as a new national 24-hour rainfall record. The flood damaged a large portion of Kauai County and portions of the City and County of Honolulu. Record flooding left several people dead in April and May of 2017 in southern Missouri and northwest Arkansas, where a study of changes in the peak stream flows since 1930 indicated consistent positive increases of up to 10 percent per year.
To locate areas in Honolulu that are susceptible to flooding, we used topographic and tidal data. First, we identified areas vulnerable to one meter of sea level rise and established a base water level in urban Honolulu, designated as the area between Pearl Harbor and Diamond Head (Figure 1).
Figure 1. Study area: urban Honolulu, Oahu
Three hazard models then simulated flooding in addition to the base water level. One produced storm surge flood depth, flow speed, and inundation areas for a Category 4 storm similar to Hurricane Iniki, which hit Kauai in 1992. Another simulated inundation based on five previous tsunamis that had impacted Hawaii. Another model was used to generate a 500-year river flooding, capturing inland flooding from heavy rainfall. Areas of flow inundation show the extent of the exposure to the flooding, while flood depth and flow speed illustrate the potential for damage.
Regional transport models, such as the Oahu Metropolitan Planning Organization (OahuMPO) transportation demand model, are typically used as forecasting tools to inform planning decisions. However, transportation demand models are not able to capture the nuances of the evacuation decision-making process. There are no evacuation travel surveys for Honolulu, so we used OahuMPO model estimates of expected demand for transportation infrastructure, the origins and destinations across motorized and non-motorized modes, and historical travel data to estimate how people might try to get away from each disaster.
Figure 2. Inundation due to sea level rise, tsunami, and hurricane storm surge, and inland surface flooding
Figure 2 shows the maximum depth of anticipated flooding from sea level rise, tsunami, and hurricane storm surge, or inland river overflow in the study area. For each cell on the map, we identified the highest possible rise among the three inundation hazards — meaning if an area would flood up to three feet in a tsunami but five feet due to river overflow, we assigned it to five feet.
The “ruling hazard,” displayed below is what we call the most critical flooding type for each cell.
Figure 3. Ruling hazards for each cell
This ruling hazard analysis identified the maximum of maximums — that is, the flood type within the grid for each of the three different hazards (hurricane storm surge, tsunami run-up, riverine flooding) that causes the highest flood level for that location. This analysis indicates the predominant threat type. While deeper flooding generally means a higher risk of damage, different communities, utilities, and governments will have different responses to flooding from a tsunami versus that of a river.
A city exposed
The worst-case flooding scenarios would have a massive negative outcome on the people, economy, infrastructure, and emergency plans of the Honolulu area. 45 percent of the population — more than 150,000 people — would find their homes directly affected. For the 32,000 business establishments in the study area, we found that our scenarios would cost those businesses $34.8 billion, representing 80 percent of Honolulu’s economy. Employment data shows that nearly 213,000 jobs, or 76 percent of the area’s workforce, are exposed to worst-case flooding.
Many critical infrastructural facilities are also located in the worst flood zones, including all major electric power, oil, port, potable water, and wastewater facilities. Four out of five of the area’s hazardous material sites will be affected by worst-case flooding, along with 85 percent of communications facilities. Most of these systems are not built for resiliency during a flood and could fail when needed most.