In late 2005 I was on New Providence Island, Bahamas, producing
a map to show which properties around the island were within the storm surge
flood zone. The northern islands of the Bahamas had been battered by 19 feet (six
meters) of storm surge in 1999 during Hurricane Floyd and flooded again in 2004
Hurricanes’ Frances and Jeanne.
While wandering around the poorer, south side of Nassau, I came across a single-story building, probably a community center or clinic, with barred windows, on which was written “Hurricane Shelter”. It was sufficiently surprising that I even took (and kept) a photo – see below, for the “Hurricane Shelter” was only two to three feet above sea level. If people gathered at this shelter as a strong hurricane approached, they would be placing themselves in mortal danger from an accompanying storm surge.
Initial reports from the Bahamas suggest that the islands of
Great Abaco and Grand Bahama have been left devastated from Major Hurricane
Dorian, evoking memories of the destruction on the eastern Caribbean island of Barbuda
in the immediate aftermath of Hurricane Irma just two years ago.
A Record-Breaking Hurricane for the Bahamas and the Atlantic
Dorian underwent an unprecedented period of rapid intensification between August 31 and September 1, that took its maximum sustained wind speed from 150 miles per hour to 185 miles per hour. No other Atlantic hurricane on record has intensified as rapidly as this from such a high initial wind speed. Dorian joins an exclusive group of Atlantic hurricanes to attain wind speeds of 185 miles per hour or greater: Allen (1980), Wilma (2005), Gilbert (1988), and the Labor Day Hurricane (1935).
Dorian maintained this intensity on September 1, and then made a series of landfalls – first across Great Abaco island, and on September 2 across Grand Bahama. In doing so, Dorian became the strongest hurricane in modern records to strike the northwestern Bahamas. As the Category 5 hurricane traversed the islands, its forward speed slowed and it became near stationary over Grand Bahama for roughly 36 hours before gradually moving northwest. Dorian’s eyewall subjected some areas of these islands to destructive wind gusts of up to 220 miles per hour (354 kilometers per hour) and catastrophic storm surge in excess of 20 feet (6 meters).
Dorian looks set to pass over the northern Bahamas in the coming days as
potentially a Category 5 major hurricane, but forecasts regarding future U.S.
impacts remain significantly uncertain, with the latest guidance providing a
twist in the tale that no one anticipated a few days ago.
Understanding the Uncertainty: A Matter of Timing
The meteorological situation that Hurricane Dorian finds itself in is as fascinating as it is uncertain. Several days ago, Florida was bracing itself for potentially its third major hurricane landfall in as many years. Now, Dorian looks more likely to make landfall in the Carolinas, or, as some models increasingly suggest, it may recurve soon enough that is misses the U.S. entirely. So, why have the forecasts been so uncertain? It’s all to do with timing.
Now that we’ve reached the halfway stage of the 2019 North Atlantic hurricane season, now feels like a good opportunity to review the season to date and look ahead to what the remainder of the season might have in store.
A Quiet Start to the Season
If you thought the Atlantic had been a little quiet through
the early summer, you’d be correct. The basin has had its quietest start since
2014. The strongest of these storms to date, Barry, made landfall near
Intercoastal City, Louisiana, on July 13 as a weak Category 1 hurricane. RMS
estimated that the insured U.S. losses from Hurricane Barry would not exceed US$500
million, inclusive of wind, storm surge, and inland flood damage, including
losses to the National Flood Insurance Program (NFIP).
Every twist and
turn of a real-time hurricane can affect global financial markets, public
safety, or government and international aid agencies that provide assistance. Within
the (re)insurance space, the ability to understand forecast track, timing, and potential
hazard and loss impacts before landfall helps entities to prepare and execute
their event response processes effectively. This includes having adequate
capital to cover claims, setting up claim centers and planning policyholder
outreach, securing and positioning adjusters in areas that are likely to be
impacted, and determining what, if any, risk can be ceded to reinsurance or
clients, the traditional approach to quantify potential impacts ahead of a
landfalling storm involves selecting similar storms from the RMS® North
Atlantic Hurricane (NAHU) stochastic event set. While this generates vital
insights that can be extracted quickly from internal databases, there are
opportunities to provide earlier and more comprehensive insights into the storm
ahead of landfall.
To date, RMS clients have also benefited from real-time analysis of hurricane events through RMS HWind Real-Time Analysis products. These observation data-based snapshots and footprints have provided the industry with a standard “ground truth” representation of tropical cyclone wind field size and intensity before, during, and following landfall effectively helping to describe what the storm is doing and what the storm has done.
Earlier this year, RMS released its latest medium-term rates (MTR) forecast for the North Atlantic hurricane basin as part of the North Atlantic Hurricane Models Version 18.1 release. Applicable over the 2019-2023 period, the Version 18.1 forecast represents an update from the previous MTR forecast issued in 2017 for the 2017-2021 period, by reflecting hurricane activity from the 2017 and 2018 seasons.
The MTR forecast provides a forward-looking estimate of the expected average annual landfall rate on a five-year horizon. Available alongside the long-term rates (LTRs) – a view of hurricane frequency based on the climatological average for the period from 1900 onwards, MTRs provide an additional perspective on expected hurricane rates on a shorter timescale. This allows RMS to adjust our view of risk according to the observed climate variability, and to combine different scientific theories on the drivers of hurricane variability over time, ultimately providing a view of landfalling hurricane risk that best represents the near-term basin conditions.
The 2019 North Atlantic hurricane season officially got underway on Saturday, June 1, and marked the start of a six-month period that runs right through to November 30. Blatantly ignoring this official start, the North Atlantic has already produced its first named storm of 2019. On May 20, Subtropical Storm Andrea formed over open water in the western Atlantic, several hundred miles south of Bermuda. It was a relatively weak and short-lived storm, lasting for less than a day before dissipating. This is the fifth consecutive year that a storm had formed ahead of the official start date of the hurricane season.
As I shared in a previous blog, storms can form at any time of year, but it is important to remember that there is no historical relationship between the date of the first named storm and the overall seasonal hurricane activity, so the early start to 2019 does not provide us with any clues as to how the season might pan out.
With the release of version 18.1 on April 22 from RMS, there is plenty to explore, validate and put into production.
Updated Insights on North Atlantic Hurricane Risk
Starting with the RMS North Atlantic Hurricane (NAHU) Models, version 18.1 (v18.1) includes updates to the long-term and medium-term event rates throughout the Atlantic Basin, historical event reconstructions from recent seasons, and hazard and line-of-business specific vulnerability enhancements informed by new data and RMS building research.
Why the Saffir-Simpson Hurricane Intensity Scale had five levels we don’t know. The digits on a hand? Better than three, but lower resolution than the dozen rungs for wind speeds or earthquake intensity? Whatever the reason it seems to work.
In the late 1960s, Herbert Saffir, a Florida building engineer, was sent by the United Nations to study the hurricane vulnerability of low-cost housing in the Caribbean. He realized something was needed to rank hurricane destructiveness. Saffir had some “Richter envy” from observing the ease with which seismologists now communicated with the public. In 1971, he contacted Robert Simpson, head of the National Hurricane Center to help link damage levels with wind speeds.
Seeing the opportunity to communicate evacuation warnings, Simpson also added details around the height of advancing storm surges. Better information was clearly needed, after the loss of life in Hurricane Camille on the Mississippi coast in 1969.
Professor Ilan Noy holds a unique ”Chair in the Economics of Disasters” at the Victoria University of Wellington, New Zealand. He has proposed in a couple of research papers that instead of counting disaster deaths and economic costs, we should report the “expected life-years” lost, not only for human casualties but also for the life-years of work that will be required to repair all the damage to buildings and infrastructure.
The idea is based on the World Health Organization’s Disability Adjusted Life Years (DALYs) lost through disease and injury (WHO 2013). The motivation is to escape from the distortion introduced by measuring the impact of global disasters in dollars, as loss from the richest countries will always dominate this metric. Noy’s proposal converts injuries into life-years lost, based on how long it takes for the injured to return to complete health, while also factoring the degree of permanent disability multiplied by its duration. This is topped up by a “welfare reduction weight” for all those exposed to a disaster. The final component of the index attempts to capture how many years of human endeavor is lost to recovering the buildings and assets destroyed in the disaster.
There is plenty to argue over in terms of how deaths, injury and damage should be combined. In particular, the assumption that additional work to rebuild a city, is the same as a shortened life, seems somewhat reductive.