5 Common Hurricane Response Plan Problems and How to Fix Them

Recent hurricane seasons have been some of the most active on record, with 21 named storms, seven hurricanes in 2021, and 30 named storms, 14 hurricanes in 2020. Looking ahead to the 2022 Atlantic hurricane season, observed and forecasted conditions, including the low probability of an El Niño, above-normal ocean heat content, and decreased low-level shear in the main development region point to another hurricane season with an above-normal number of named storms.

As tropical storms and hurricanes are one of the leading causes of death and destruction from weather disasters, businesses have learned that disaster management and hurricane response plans are essential. According to the NOAA Office for Coastal Management, tropical cyclones have generated over $900 billion in damage, with an average cost of almost $21.5 billion per event, and led to over 6,500 deaths between 1980 and 2020.

According to the National Hurricane Center, 59% of all business disruptions are caused by weather. For any industry, companies that plan for a disaster have the best chance of recovering and resuming their business. Fortunately, many businesses are taking the necessary steps to protect themselves and their employees by developing hurricane response plans. Often these plans have considerable problems that may significantly limit a businesses' ability to ensure personnel safety and maintain business continuity during hurricane seasons.

Hurricanes and tropical storms are nothing new, and the U.S. has a long history of dealing with their devastating impacts. Having worked with hundreds of businesses for four decades, both before and after a hurricane has hit, these are the five most common problems seen in most hurricane response plans along with the strategies that might help you solve them.

1. Incorrectly Interpreting Cones of Uncertainty

One of the most common problems with hurricane response plans is using the "cone of uncertainty," or error cone, as a predictor of whether a storm will impact a specific location. Most people are familiar with the error cone around the storm track, and it is often interpreted as being a zone where the storm will cause damage. However, the cone of uncertainty has little to say about the extent of the hurricane's impact.

The cone of uncertainty is a graphical representation of forecast error added to tropical cyclone forecast track maps. The cone is not based on the meteorology for a given storm but on forecast error rates of storms from the previous five years. This means that the center of a storm will track within the predetermined cone of uncertainty 75% of the time based on reviews of forecasts from the past five years. The cone size is defined at the beginning of each season. It's always the same size for every storm and every forecast. At the beginning of the forecast period, the cone is typically narrow to reflect low forecast error. As the forecast extends into the future with progressively more error, the cone becomes wider and increases in size.

It is a common misunderstanding that storm surges, rainfall, and winds are confined to the area within the confines of the cone. The effects of a tropical storm can reach hundreds of miles outside the cone. Over time, with increased data and more accurate predictions, the cone of uncertainty has decreased in size. The cone does not directly relate to storm size, meaning that just because a location sits outside the error cone does not suggest it will be spared by the storm.

The limitations and frequent misconceptions of the error cone are what drove me and a group of meteorologists at StormGeo to develop the Threatened Regions from Active Cyclones (TRAC) model. TRAC analyzes over 100 predictive models to map the probability that the location will receive tropical-storm-force winds resulting in a more accurate prediction of areas at risk. By graphically representing the probabilities of the storm track, TRAC also is better able to illustrate forecast uncertainty.

Businesses looking to improve their hurricane response plans should rethink triggers based on whether or not the monitored location lies within the error cone. A better trigger would be to reference a model like TRAC and set triggers based on the probability of wind impact.

2. Misunderstanding the Saffir-Simpson Hurricane Wind Scale

Another problem often seen in hurricane response plans is the incorrect use of the Saffir-Simpson Hurricane Wind Scale. The Saffir-Simpson scale was introduced to the public in 1973 by civil engineer Herbert Saffir and meteorologist Robert Simpson to estimate the level of damage winds would cause to buildings. The well-known scale organizes hurricanes into five categories based on the storm's maximum average wind speed over a one-minute time span.

Many companies base their evacuation decisions on a hurricane's predicted Saffir-Sampson category because of its easy-to-understand rating system and its long history with the public. However, the scale is narrow in its focus. It categorizes storms solely by their wind speed without saying anything about the size of the storm and potential storm surge inundation, which typically is associated with most injuries and deaths relating to tropical cyclones.

For example, a response plan may require an evacuation of a facility only after a hurricane reaches Category 3 or higher, leaving crews to ride out Category 1 and 2 hurricanes. However, in some cases, a Category 2 hurricane could produce a more significant storm surge than a Category 4 hurricane based on the size of the wind field. To determine the true damage potential of a storm, both the maximum wind speed and the size of the wind field must be considered.

In addressing the Saffir-Simpson scale's limitations, my colleagues and I developed a new rating system that better estimates a hurricane's potential for damage: the Hurricane Severity Index (HSI). This index is based on a fifty-point system where half of the points come from the size of the storm's wind field and the other half from the storm's max sustained wind speeds. By combining these two factors, the Hurricane Severity Index is better equipped than the Saffir-Simpson scale to predict the overall severity of a storm.

3. Poorly Defined Proximity Triggers

Proximity triggers are another common element in hurricane response plans. These triggers use the storm’s location to activate a specific action in a hurricane response plan. When a storm comes within a certain proximity to a location, this 'trigger' will instigate response actions. The use of a proximity trigger may cause your business to take unnecessary steps in your plan, which costs time and money.

For example, if we look at Hurricane Wilma, its proximity to the upper Texas coast pushed several businesses to activate their hurricane response plans. However, meteorologists were reasonably confident that there was a jet stream blowing across the Gulf from the southwest to the northeast that would drive the storm across Florida and eliminate the threat to businesses located on the upper Texas coast.

The problem with proximity triggers is that proximity does not necessarily correlate with the threat level for a specific location. Just because a hurricane is “x” miles from a location does not mean it poses any threat at all. In addition, proximity triggers do not take into consideration the hurricane’s changing forward speed or its size.

When developing response plan triggers, companies need to incorporate more than just the current location of the storm. A strong trigger incorporates the forecast time of wind arrival, probability of wind impacts, and the worst-case estimated time of arrival. This ensures the response plan accounts for changes to the track, speed, and intensity of the storm.

4. Not Taking Potential Flooding into Consideration

The U.S. has witnessed several devastating flooding situations in recent years. Hurricane Katrina led to massive floodings in New Orleans. Hurricane Harvey unleashed the worst flooding in the history of Houston.

Tropical storms and hurricanes often cause flash flooding from heavy rains, especially inland from the coast, and hurricane response plans should consider this fact. Many response plans also overlook the threat to flooding from tropical storms. Tropical Storms Allison (2001) and Beta (2020) brought 40” and 15” of rain, respectively, to the Houston area, and Hurricane Henri made landfall as a tropical storm but still caused widespread flooding across the northeast.

Rainfall is independent of the intensity of the system. A tropical depression or remnant low-pressure area could produce heavier rain than a strong hurricane. Speed of movement is the key. A slower-moving system drops heavier rainfall. Although it is difficult to predict precisely where rain might fall, it is possible to determine if a storm will stall and cause flooding due to its delayed movement through an area.

5. Basing Your Action Timeline on a Single Deterministic Forecast

Finally, the biggest problem we see with hurricane response plans is the reliance on just one deterministic forecast. The deterministic forecast is what the meteorologist thinks is the most likely track and intensity. Basing your hurricane response plan on one forecast is risky, especially if the forecast is made to the general public about a wide region. This forces you to assume that the impacts on your location will be consistent with what the rest of your region sees. Businesses basing their decisions to take action only if a particular wind speed is projected to impact their location on that one area-wide deterministic forecast may be making a critical mistake. There may be considerable uncertainty in the forecast. Forecast uncertainty needs to be considered when deciding whether or not your business needs to take action.

Most hurricane plans are initiated by a proximity forecast, mapping out actions to fixed hourly times. For example, a facility might decide to take a specific action 96 hours “out” before the hurricane arrives. The problem is that the meaning of " out" is rarely adequately defined and subsequently becomes subjective and inaccurate. Different decision-makers could be defining arrival as the first recorded tropical-storm-force winds or be based on the hurricane eye. Without being a hurricane expert, it is tough to determine when you are within, for example, 96 hours of being impacted by a hurricane.

Hurricanes are known to change rapidly, making flexibility key to managing your response. Instead of relying on deterministic forecasts and static goalposts, businesses should consider taking forecast uncertainty into account by generating a timeline for the evolution of a storm and calculating the likelihood of impact on a specific location. This includes pairing the storm's projected track and worst-case arrival times with the probability that the storm will impact the area based on the projected size of the wind field. By considering forecast uncertainty, previously subjective triggers can be transformed into objective triggers that more accurately tells you when certain actions should be taken and when future actions will need to be taken.

Knowing these common shortfalls in hurricane response planning, businesses using any of these five parameters as triggers should review their existing plans, procedures, and protocols. Are you using the right tools and models to prepare for hurricane impact? Are you basing your decisions on accurate and reliable data vetted by a team of experienced meteorologists? The benefits are considerable: If you plan accurately, you can ensure safety and success through the upcoming hurricane season.

Chris Hebert is StormGeo's lead hurricane forecaster. He manages the TropicsWatch team, helping clients visualize and better understand the overall risk from a hurricane or tropical storm. Along with his team, he provides tropical expertise, forecasts, and services for all ocean basins to all global StormGeo clients at any operating location.

Originally published in the Disaster Recovery Journal.