You ever watch a tornado tear through the countryside on TV and think, “Wow, that’s terrifying… but also kinda fascinating? I’d like to see one in real life.” No? Just me? Okay, moving on.
For my Foundations of Data Science class, I analyzed data on extreme weather because what’s more fun than tracking Mother Nature’s rage-induced spirals of destruction known as tornados? Also, my professor made it a mandatory assignment which I needed to accomplish; otherwise, I’d never end up graduating!
Step 1: Getting the data
Where does one find tornado data? Turns out, the U.S. government has been tracking tornados for decades. Agencies like NOAA (National Oceanic and Atmospheric Administration) keep detailed records of every twister that has ever yeeted a cow across a field.
So here I am battling messy CSV files with weirdly labeled columns.
Step 2: Craft beautiful graphs using DataWrapper
Once I had the data, it was time to visualize it. Because nothing says “I understand things” like beautifully crafted line charts of agony.
Step 3: Realize your line charts were boring and make a symbol map
This symbol map was much more enticing, beautiful, and also terrifying if you live in the midwest.
Step 4: Finding the Patterns
Now for the fun part—analyzing trends:
- Tornadoes love springtime, which makes sense because who doesn’t love the springtime?
- The Southeast has more to worry about, meaning Florida and Georgia are even more terrible than before.
Step 5: Make an entirely different graph between Georgia vs Oregon because your professor demands it!
Final thoughts: What I learned
- Data is cool, but tornadoes are terrifying, and I will most definitely never move to the Midwest.
- DataWrapper and Voyant make you seem like you know what you’re discussing.
- If you see a cow flying past your window, it’s time to take cover.
TL;DR
I used data science to track tornadoes, and now I have trust issues with the sky.
