This project was carried out as part of the Data Analytics course at the Ironhack programming school. The developed work falls within the scope of climate change in Continental Portugal, more specifically, the change of the variables associated with temperature over time.
This analysis used Reanalysis Data that included minimum, average and maximum daily temperatures in 1012 locations in Continental Portugal between 1980 and 2021. The goal of the investigation was to analyze the evolution of temperature over that period in several dimensions. For that purpose, the total period of 42 years was divided into two consecutive shorter periods of 21 years to be possible to compare them.
Among the tasks that constituted the project, some of them are:
- An evolution analysis of the average daily temperature per period, and the average daily thermal amplitude per year.
- A cluster analysis using the k-means method, which allowed grouping all months of the year into 4 different classes. That analysis was made in each period and according to the characteristics presented.
- A quantitative and qualitative comparative analysis of extreme maximum temperatures (above the 90% percentile), which made it possible to determine the difference between the number of occurrences of these extreme events and their termic intensity in each location and period. In order to verify the significance of the increase in the intensity of occurrences, a hypothesis test was carried out with a significance level of 5%.
Finally, all data processing and analyzes were performed in Python, and the results were presented in Tableau. The respective Notebook and Dashboard are available in this repository.
(The database used in this analysis exceeds GitHub's file size limit of 100 MB. In following link it is possible to access the csv documents that allow the user to run the code: https://drive.google.com/drive/folders/1O5QWDHgI5I_A-qVXerTOskWlQ5LCJrkS?usp=drive_link)