One of the challenges in FreeeCodeCamp visualization course was creating a heatmap about global temperature over the last 200 years. There was a visible pattern showing that the recent 50 years are hotter than the average temperature from the two recent centuries. Inspired by that chart, I wanted to check if a similar pattern is visible for Poland in recent years. I didn’t aim to confirm/reject any thesis but to check if the climate crisis is so significant in Poland, that human eye can easily detect on the chart.
Dataset & preprocessing
The data with temperatures in Poland comes from IMGW (governmental) public dataset. Records with the condition
(station_name == "WROCŁAW" and year <= 1965)
were excluded due to suspected temperature (0°C every month). To get the same reference point for each month, the data is represented as Z-score (number of standard deviations from the mean). All the data (mean temperature for the given month, the mean temperature for the month over years and z-score) are visible on the tooltip:
How to interpret it? The average temperature in December 1971 was 1.1°C. The mean temperature for December (1966-2019) was -1°C. The temperature in the observed month was 0.97 standard deviations from the December mean (-1°C).
The chart below shows Z-score for 6 cities in Poland: Warszawa, Lesko, Wrocław, Suwałki, Szczecin, Zakopane, each of them represents a different region. There are among them both the biggest cities in the country (Warszawa, Wrocław, Szczecin; >400k citizens), the medium (Suwałki, Zakopane; ~50k citizens) and the small one (Lesko, ~5k citizens).
All the cities have a similar pattern to the chart from FreeCodeCamp, namely, most of the monthly temperatures in the XXI century are above the average from the recent 50 years.
The preprocessing and frontend code can be found at my Github.