Hurricane season in the Atlantic Ocean runs annually from June 1 to November 30. The number of hurricanes that develop each year varies, though meteorologists endeavor to make predictions about the number and intensity of storms every year. For example, predictions from the National Ocean and Atmospheric Administration (NOAA) for the 2025 hurricane season were for six to ten hurricanes of which three to five were expected to become major hurricanes [1]. The Colorado State University Tropical Cyclones, Radar, Atmospheric Modeling and Software Team also had similar predictions of eight hurricanes with three being major [2]. Hurricanes are categorized according to the Saffir-Simpson Hurricane Wind Scale, which rates them from category 1 to 5 based on sustained wind speed. A tropical storm is named by NOAA when its winds reach 39 miles per hour (mph). It becomes a category 1 hurricane at 74 mph and a major hurricane (category 3) at winds of 111 mph. The 2025 hurricane season resulted in five hurricanes, four of which were major hurricanes. Three of the four reached category 5 (winds 157 mph or greater) and the fourth reached category 4 (winds 130 – 156 mph). None of the hurricanes in the 2025 season made landfall in the United States. There was one tropical storm, named Chantal, that did make landfall and caused approximately $500 million USD in damages. [2] There is considerable history of tracking hurricane landfall. NOAA records go back to 1851, however, there is lower confidence in numbers before 1901 due to regions of the U.S. coastline that were relatively unpopulated in that period. From 1901 to 2024 there have been 213 hurricanes to make landfall in the U.S. Seventy-two of those storms were major hurricanes and only four were category 5. From 2021 to 2024, nine hurricanes have made landfall in the U.S. and one of them was category 3 and the other three were category 4 [4].
The impact of a hurricane making landfall is a well-known phenomenon. The aftermath of a hurricane can include damage to property from winds and storm surge, extended power outages and loss of other utilities, and large portions of a community evacuated and unable to return home or stranded in place due to obstacles to transportation. Those who live in “hurricane country” have learned how to be prepared – what to have on hand or buy before the storm arrives, where to get storm updates, and what their plan is in the event of an evacuation order. A hurricane is a somewhat unique natural disaster because those impacted have often watched the storm gradually develop and strengthen as it progresses towards possible landfall locations. Someone in the path of a hurricane knows it is coming but not how bad it will be until after it has arrived or moved through. This progression gives communities in the potential path of a storm the opportunity to prepare for the storm’s arrival. There are a number of ways communities prepare for hurricanes. One of the key preparations is for individuals to make sure that their vehicles are fully fueled in the event they have to evacuate. For those who plan to shelter in place and have resources like generators, they need to have fuel on hand in the event of power outages, which are quite common following a hurricane.
Extreme weather events significantly impact the supply and demand of gas leading to short and long term fluctuations in prices. Impending extreme weather can disrupt the infrastructure that mines the gas along with the transport of gas, limiting the supply, leading to an increase in prices. Extreme weather also alters consumer behavior. Consumers tend to stock up when warned of an incoming storm because they may require more gas for evacuating or for running generators. Historically, consumers have experienced dramatic rises in gas prices due to severe hurricanes. Directly after Hurricane Katrina, U.S. consumers saw a 46 cent increase in average gas price in just one week [5]. This hurricane not only impacted gas prices among communities that were struck by the hurricane, but across the nation as well. Approximately 20 percent of U.S. gas production takes place in the Gulf, so nationwide gas prices are especially vulnerable to change with hurricanes that occur in the Gulf [5]. Being able to predict the changes in gas prices leading up to extreme weather helps consumer safety by ensuring they can evacuate or run essential equipment.
Those impacted by hurricanes and other large scale weather events are not limited to individuals in the path of the event. Major industries, including financial, healthcare, insurance, and government sectors are severely impacted by hurricanes, either due to direct physical ramifications or due to consumer expectation. Specifically, the Gulf of Mexico is a hub of industry for the natural gas industry. Thus, there is volatility in the nationwide wholesale supply chain for offshore petroleum when a hurricane is present in the Gulf of Mexico or makes landfall. Due to the potential that aggregated consumer expectations have on markets, even when hurricanes do not make landfall, the commodity markets relating to utilities, specifically oil and gas, are impacted. However, the stakeholders most susceptible to large scale weather events are arguably individuals and households because of real physical danger posed by storms and their vulnerability to financial shock.
Increased understanding of the impact of forecast hurricanes on retail gas has the ability to reduce financial impact to households in the path of the hurricane. Real people are impacted when gas prices are inflated; therefore, real people can benefit from a robust analysis to support modeling of weather-driven changes to gas prices.
To understand the impact of major storm events in the Atlantic ocean on retail gas prices, we will utilize the AAA’s fuel price website, which is a collection of metropolitan gas prices. This is a reliable source that houses data regarding the daily average gas prices of diesel, low, regular, and premium fuel for major metropolitan areas in the United States, most relevantly, major cities in Florida, both in the peninsula and panhandle.
We will use the wayback machine in conjunction with this AAA source to parse historical data. We will scrape historical data to generate a data frame that holds information on the average price a week before the Internet Archive grab. This will allow us to understand the gas price history using time series techniques.
Additional historical gas price data, available from the U.S. Energy Information Administration, will be used to compare to collected AAA data as well as to provide reference data for gas prices in other parts of the United States.
Tropical storm and hurricane data from the National Oceanic and Atmospheric Administration (NOAA) will be used to evaluate the progression, status, and position of extreme tropical weather systems. The NOAA data sets include a variety of tropical weather characteristic fields which will be examined to understand relationships and identify indicators with the potential to correlate to gas prices. Additionally, weather data will be used to investigate the feasibility of isolating tropical weather impacts from the impacts of other factors on gas prices.
To answer the causal question regarding the impact of extreme weather events on gas prices, a staggered rollout model could corroborate with the natural pattern of a hurricane so we could analyze the difference between an area hit by the hurricane and one that is down the path, not yet fully impacted, while holding time constant.
Furthermore, we can utilize a basic linear regression or a Ridge Regression to understand the percent change in gas along with hurricane indicators. Regressing gas price on natural events with covariates including time, price of other goods, and macroeconomics indicators could allow us to understand underlying relationships and statistical significance.
Finally, we will employ a clustering model such as k-means or a different unsupervised model to give robust recommendations to stakeholders regarding the optimal time to purchase gas.
Hurricanes are monitored and tracked in great detail, providing an abundance of data that can be used to learn more about the evolution of a hurricane. NOAA utilizes three numeric models–HWRF, HWRF-B, and FV3 HAFS–to forecast hurricane activity, storm strength, and hurricane structure. They also implement statistical dynamic modeling (SHIPS) which estimates the probability of intensification of a tropical storm [6]. Once a storm is detected, wind speed and direction of travel of the storm are monitored to characterize the storm as defined by the Saffir-Simpson Hurricane Scale [3]
The current gaps in the literature regard the impact of hurricanes on consumer behaviour. Although economic theory proposes that raising prices of essential items could allow for efficient allocation of resources, through anti-hoarding, anti-price gouging laws have been implemented in Florida to keep companies from dramatically increasing prices when a state of emergency is in place [7]. Furthermore, while there exists research on the change in the commodity price of gasoline and natural gas, viable consumer advice regarding price timelines for retail petroleum, in terms of time till landfall, for essential goods is also not well established [8].
Existing solutions for households is limited to individuals monitoring news sources and social media to understand when to acquire gas and other essential goods at the best prices. This solution is limited by the presence of misinformation in social media and timeliness of news releases with respect to real time gas price changes. The efficient utilization of the integration of economic indicators, historic gas prices, and published weather data for a predictive model serves as a new multidisciplinary solution for households other stakeholders.
[1] NOAA. 2025. NOAA predicts above-normal 2025 Atlantic hurricane season. Retrieved January 28, 2026 from https://www.noaa.gov/news-release/noaa-predicts-above-normal-2025-atlantic-hurricane-season
[2] Colorado State University. Colorado State University Seasonal Hurricane Forecasting. Retrieved January 28, 2026 from https://tropical.colostate.edu/forecasting.html
[3] National Weather Service. Saffir-Simpson Hurricane Wind Scale. Retrieved January 28, 2026 from https://www.weather.gov/mfl/saffirsimpson
[4] National Weather Service. U.S. Hurricane Strikes by Decade. Retrieved January 28, 2026 from https://www.nhc.noaa.gov/pastdec.shtml
[5] Rob Grunewald. 2005. Gasoline prices climb in response to hurricanes. Federal Reserve Bank of Minneapolis. Retrieved February 1, 2026 from https://www.minneapolisfed.org/article/2005/gasoline-prices-climb-in-response-to-hurricanes
[6] NOAA. 2025. AOML's Hurricane Modeling & Prediction Program. Retrieved February 2, 2026 from https://www.aoml.noaa.gov/hurricane-modeling-prediction/
[7] State of Florida. 2026. Price Gouging Statute. Retrieved February 2, 2026 from https://www.myfloridalegal.com/price-gouging-hotline/price-gouging-statute
[8] US Energy Information Administration. 2024. Gasoline explained. Retrieved February 3, 2026 from https://www.eia.gov/energyexplained/gasoline/price-fluctuations.php
[9] U.S. Energy Information Administration. 2024. Florida's retail gasoline price stays stable after Hurricane Milton despite shortages. Retrieved February 6, 2026 from https://www.eia.gov/todayinenergy/detail.php?id=63466
[10] NOAA. 2025. NOAA Historical Hurricane Tracks. Retrieved February 6, 2026 from https://coast.noaa.gov/hurricanes/#map=4/32/-80
[11] Marine Cadastre. 2025. Offshore Oil and Gas Platforms. Retrieved February 6, 2026 from https://hub.marinecadastre.gov/datasets/offshore-oil-and-gas-platforms/explore?location=26.619305%2C-88.563475%2C6