Join the Alteryx Community’s Maveryx Summer Cup event! Compete, network with others, and earn your gold through a series of challenges from July 24th to August 11th. Learn more about the event here.

Data Science

Machine learning & data science for beginners and experts alike.
rachel_k
Alteryx
Alteryx

2024 Atlantic hurricane season is approaching quickly.  Unusually high water surface temperatures and potential transition to La Nina conditions later this year both suggest a potentially active hurricane season in 2024 (reference: Ominous Signals For The 2024 Hurricane Season (forbes.com)).  

 

The National Oceanic and Atmospheric Administration (NOAA) estimates that hurricanes have caused $1.3 trillion in damage since 1980 (through August 2023), an average cost of $22.8 billion (about $70 per person in the US) per event. The impacts on businesses are a major factor: property damage, disruption to supply chains, loss of customers, and insurance shortfalls. Hurricane insurance may not fully cover these problems, let alone potential workforce loss during and after the event.

 

Emergency planning can help companies minimize losses. Key things businesses want to know with an approaching storm include:

 

  1. Where is it going?
  2. How bad will it be?
  3. How might that affect my facilities?

 

For now, let’s focus on questions 1 and 3 – where is the storm going, and which of your facilities could be affected? There’s no crystal ball, but there is Alteryx. Let’s start by understanding a little more about hurricane forecasting and then how Alteryx can help you customize that forecast for your organization.

 

Hurricane Forecasting

 

Hurricane forecasting is tricky, even for the experts at NOAA – the home of the National Hurricane Center (NHC). Global winds, pressure fronts, water temperature, air temperature, storm movement over land, and more can combine to alter the path and intensity of a storm. That’s why it’s hard to accurately forecast a storm a week– or even a few days in advance. There are too many unknowns. Often, changes happen in the few hours before a storm makes landfall. The NHC updates its forecasts every six hours based on the most recent weather and ocean observations.

 

The forecast includes a storm track – a single line with the predicted latitude and longitude of the eye (center) of the storm in 12-hour intervals. Because forecasting is imprecise, a cone is projected around the track, like the cone outlined below. The cone is based on historical predictions and observations. There is a 66% likelihood that the eye of the storm will be within that region. That also means there is a one-third chance that the storm will not be in that cone. The cone gets bigger as the time interval of the forecast increases – again, that’s because forecasting is imprecise. The further out the prediction, the more uncertainty there is.

 

image001.png

 NOAA-NHC hurricane track forecast for Hurricane Idalia. Circles represent the predicted location of the eye of the hurricane, and the shaded/dotted cones represent the area with ~66% likelihood of the path of the eye.

 

The cone gives an idea of where the storm is most likely to hit, but it isn’t perfect for emergency planning. Pretend you’re an insurance company covering all of Florida’s public libraries (shown below). 

 

image002.png

 

Looking at the NHC forecast, it’s easy to find some locations clearly within the prediction code, such as the Reddick Public Library, in red. It’s harder for libraries near the edge of the cone.  If your insurance company wants to be more careful and identify libraries in a larger 80% likelihood cone, it’s even harder to rule-in at-risk locations based on the NHC cone image.

 

Forecasting With the Power of Alteryx

 

The Alteryx Professional Services data science team has created an end-to-end process to assist in assessing facilities risk:

  • First, an Alteryx workflow is used to go to the National Hurricane Center’s website and download the most recent forecast.
  • Then, a series of processing steps are applied to pull key information from the forecast – in this case, the expected latitude and longitude of the storm over time – and then calculate uncertainty cones.
  • Lastly, maps are created that overlay the predicted storm track and uncertainty cone with a list of your facilities. A customized report identifies facilities in the potential storm path.
  • To maximize enhanced decision making, schedule the workflow to run on Alteryx Server each time hurricane forecasts are updated (every 6 hours) for the most up-to-date information. The email report feature can keep key decision-makers updated as new data is available.

 

Go from best-guessing to informed decision making in minutes by rapidly identifying at-risk facilities:

 

image003.png

 

The workflow (see below) is simple and only requires users to update information in the two yellow containers. In the upper yellow container, update the Text Input tool with the current named storm (just the name, i.e., Franklin, Idalia, Katia). In the lower Text Input tool, replace the list of Florida public libraries with a list of your organization’s facilities, including columns for the facility name, address, city, state, and zip code. Then press Run, and a customized report identifying your at-risk facilities will be created.

 

Note: This workflow (on the Gallery here) will require license, download, and installation of a spatial data pack (for geocoding addresses to geospatial points) and the Alteryx predictive tools, both available from the Alteryx Download Portal.

 

image004.png

 

Getting More Conservative With the Estimates

 

To create a more conservative storm track estimate (such as the region covering 80% likelihood of the storm track), remove the two connections from the 66th percentile formula tool going downward to the Spatial Match and Report Map tool, and create new connections to those tools from the 80th percentile formula tool.

 

image005.png

 

Want More? Ask Professional Services

 

The professional services team has experts in hurricane analytics, geospatial tools, visualization, and supply chain management.  We can help design and build a custom solution for you.  For example, we can incorporate regions with official watches/warnings, predicted windspeeds, alternate forecasting models (such as HWRF (Hurricane Weather Research and Forecasting) or COAMPS (Coupled Ocean Atmosphere Mesoscale Prediction System)), population impacted, impact to your supply chain, flooding, and more.  Click here to get in touch with our team!

 

About Alteryx Professional Services

 

The Alteryx Professional Services team is a group of trusted advisors who can assist you as you define, develop, and execute your vision for the Alteryx platform.  We accelerate your analytics capabilities through deployment and migration, creating process control and governance frameworks, building automation efficiencies, or developing custom analytic solutions. Get in touch if you want help on your analytic journey!