The goal while creating visualizations was to narrow the focus on the time periods and specific storms we will train the model on. We also wanted to gain context on gas price trends for our narrowed time period and metros.
2005 and 2020 saw the most hurricanes in the Atlantic from 2000-2025, while 2009 and 2013 saw the least. There is no clear trend of frequency of hurricanes in the last twenty-five years. We next wanted to see if the strength of hurricanes has shown any interesting trends over this same time period.
It appears that category 1 hurricanes occur most frequently while category 5 hurricanes are the least frequent.
Due to the data available for gas prices, we chose to focus our study on the 2022-2025 hurricane seasons. Looking at the two charts above, this time period has no significant difference from the rest of the hurricane data set, so we are confident that this time span will provide adequate results.
To gain more insights on the characteristics of the hurricanes in this time span, the team investigated the distributions of storm and wind speed. The data on storm and wind speed includes data leading up to hurricane categorization and after.
Most hurricanes range in speed from five to fifteen knots. The distribution is right skewed with outliers reaching beyond fifty knots. Outliers in this case were kept, as the faster storms could impact gas prices more than the slower.
Hurricanes are less frequent at higher wind speeds. Similarly, category 5 hurricanes are the most infrequent since they require much more energy than weaker storms.
Gas prices by metro follow the same trends in Florida with West Palm Beach consistently having the most expensive gas and Pensacola having the cheapest. Prices climbed to almost $5.00 per gallon by the middle of 2022, then returned to about $3.50 dollars per gallon by the end of the year. Starting in 2024, prices start to fluctuate much more than previous years, but the are small (less than fifty cents).
The boxplots show the differences in distributions of gas prices from 2022 to 2025. 2022 had the most variance and appears right skewed, while the other years had much less variation, and appear to be relatively symmetric.
The following charts are from the user-input filter. We filtered to Hurricane Idalia in 2023 as an example.
The first visual produced by the user filter shows the storm's path and its storm category as it traveled. The title also provides the dates of the storm. This allows the team to identify specific storms that traveled directly through Florida and the gulf.
Wind speed during Idalia quickly increased in the first few days, reaching category 4 level before dropping back below category 1 level for the rest of the storm. This gives us context that the peak duration of the storm was less than two days.
This chart shows the gas prices is different metros leading up to and after the storm. The section in blue is the exact time period of the storm. During this storm, gas prices slightly decreased, but no significant change was observed.
The final charts produced by the user-input filter show the distance to landfall of each metro for the specified hurricane and gas prices during that same time period. One example is shown below.
As stated before, gas prices during Hurricane Idalia were relatively steady. Having this pair of charts for each metro helps us identify how gas prices respond to the storm's landfall if at all.
Hurricane Idalia's (2023) distance from the metropolitan area of Jacksonville, Florida with respect to the average regular gas price. This allows us to understand, from a temporal standpoint, how storm measurements such as the wind speed (2nd chart) and storm speed (3rd chart) changed.