Using Power BI to Delve into Climate Change

I don’t know why the climate change deniers insist on denying what is going on. Tropical storm Imelda ranks right up there as the 7th wettest storm in history. Houston/Southeast Texas had Harvey two years ago and that is ranked as number 1 and is regarded as a 1000-year storm. Imelda then hits Southeast Texas again and I read that it might be categorized as another 1000-year storm. So Southeast Texas have had two 1000-year storms in 3 years. And that’s coming on the heels of hurricane Dorian, a category 5 hurricane that stalled over the Bahamas for two days and just blasted it to pieces.

The year 2017 saw Harvey slowly stroll up the Gulf Coast inundating Southeast Texas with four days of heavy rain, creating intense flooding. Maria, cat 4, ravaged Puerto Rico which still haven’t recovered today and Irma, another cat 4, travelled up Florida, leaving deaths and ruin.

In 2018 Florence flooded North Carolina and the surrounding states and hurricane Michael, a cat five, destroyed the Florida panhandle.

In 2019, Dorian, probably the strongest cat 5 ever in the US, stalled over the Bahamas. And Imelda, a tropical storm, revisited Southeast Texas as Harvey 2, bringing torrential rains and biblical flooding.

And I’m not even delving into the flooding the Midwest went through this past summer.

So, in three years we have had very intense storms and we still have climate denial? What more proof is needed? The administration is rolling back regulations protecting our air and water and regulations reducing the harmful effects of carbon dioxide. Why? What’s wrong with these people? Even the car companies are siding with California to keep the more stringent pollution rules in place.

Being interested in seeing what climate change data there is out there, especially data for the lay person like me, I thought I would try to do some research on the data and play around with them in Excel/Power BI, just to see what it tells me.

---

First Hurdle: finding data

First hurdle is finding the data, those offered for free, and grabbing a hold of them. I find the best bet are the government data, or sometimes Wikipedia in a limited fashion. The first data I found is the global land-ocean data and that data definitely showed me the upward trend of temperatures. This data came from NASA and is kind of limited in that it just showed the variance from the global average for 1951 – 1980. There is not much you can get out of this data – at least not by me.

The next set of data was the temperatures for each state. This data came from the NOAA and was a bit problematic, first in understanding how the data is pulled, then in actually pulling the data. The data had to be manually run for each state and I’m not sure I fully understand what it is I am pulling. I had set a “12-month average for December” and I’m not sure if I’m looking at an average for the year rolling from January through December or some kind of average December temperature. For the time being, this is a preliminary pull just to see what I’m getting. Later on, I will dig into trying to understand what I’m getting and to see if there is an API and whether I can actually figure out how to use the API.

This set of data, while I’m not sure I fully understand, has some interesting possibilities. I can actually set up Power BI to filter on each state so I can see what is going on for each state. With the exception of Alabama and Mississippi, it looks like temperature has been increasing since 1895.

---

Second Hurdle: making sense of data

Next I delved into hurricane data and here I find the data to be not so clear. Temperature data shows a clear rising temperature trend but hurricane data is not so clear cut. I first pulled data from Wikipedia which I think I will be able to automatically pull into Power BI in the future because it looks like I was able to set up an automatic pull in of the data when I do “get data” in Excel. I have to test this theory though. If it is true that I can pull in data into Excel through “get data”, then it stands to reason I ought to be able to automatically pull in data through Power BI.

With that said, this Wikipedia data shows only those Atlantic/Caribbean hurricanes that hit land. No tropical storms were included and tropical cyclones were excluded from the graphs. Once I pulled a count of all hurricanes hitting land, on a year by year basis, no clear discernable trend appeared. I would say frequency of hurricanes is NOT increasing and scientists have pretty much been saying that. The Wikipedia data dovetails with everything I’ve read.

The Wiki data is kind of fun to play with because you can do some kind of exploring. I tried exploring each category type hurricane but found no trends whatsoever, even for the hurricane 4 and 5, which kind of surprised me because I thought those storms would be increasing.

I did do a version to see which months had the most number of storms and August and September are pretty busy months followed by October.

But the really interesting set of data came from NOAA which tracks the path of each Atlantic storm, wave, depression, tropical storm or hurricane, since 1851. This data will be the topic of the next post because I think it has the potential to be meaty.