It’s getting hot in here: A climate analysis in R

With nearly one hundred bushfires still ravaging New South Wales, Victoria, South Australia and Tasmania, the Bureau of Meteorology has reported frequent record maximum temperatures from across the nation over the past 3 months.

I had been focusing on my local QLD figures, recorded at the Brisbane CBD station. My aim was to determine just how unprecedented the 2019 mean monthly maximum temperature profile had been.

I started by looking at the Brisbane story. As seen below, it’s been the hottest November on record.

However, as I am moving to Melbourne in 10 days time, I thought I would compare just how unprecedented the Melbourne 2019 mean maximum temperatures have been compared to the past 20 years. Winter temperatures are at around mean for the past twenty years, with summer temperatures well above average.

You can see where this is leading. I quickly wanted to compare the 2019 temperature profiles against the past 20 years of data for every capital city, and every regional centre. This was not a process I was willing to do manually.

This is where R comes in.

I’ve put an R script together to pull the long form climate record from the Australian Bureau of Meteorology and plot the latest mean monthly maximum temperature data against the historical record since 2000. There are thousands of weather stations in Australia, some with even longer complete temperature records.

The resulting charts are a replica of those completed through the manual process above.

This analysis can be repeated for any weather station in Australia, for example, here’s the story for Alice Springs, NT.

I’ve made the code open source via my github. Let me know if you find any bugs, or have suggestions for future improvements.