Visualisation of global temperature change using Tableau
In this post, the power of Tableau is fully demonstrated by providing analysis and data visualisation on the topic of global warming.
Based on the dataset provided by NASA Goddard Institue for Space Studies, the average sampling temperatures on different hemispheres are collected for each month within the time interval from 1880 to 2020.
In the visualisation, the following topics are discussed:
- Overall trend of temperature change across years
- Difference in temperature change in months
1. Visualisation on Temperature Change Across Years.
In the dataset, using the Date(unit of the month) as columns, and using the difference between the temperature measured in that month and the average temperature between 1951–1980 as row value, we can see that the overall trend of temperature change from 1880 to 2020.
If we further divide the above diagram by different hemispheres, interestingly, we can observe that the north spine currently has obviously higher rate of temperature change.
To be able to furthermore accurately measure the change rate, we take the moving average of the temperature difference between the year and the mean average between 1951–1980 for the past 10 years w.r.t the plotting year. We can see that the moving average in recent years is as high as 2.519 F.
So now we understand the overall trend in temperature change across years, let’s dive further to compare and analyze the trends in each month.
2. Visualisation on Temperature Change in Months.
If we dividied the overall trend into different months, it is not hard to see the trend in temperature change for each month.
Interestingly, if we want to compare the difference in magniture of increase in the temperature, we can draw a cycle plot and clear we will see that Janurary, Feburary and March has the highest amount of increase in temperature as compared to the rest of months.
Inspite of this, when we draw the heatmap, we can still clearly figure out that the overall trend in temperature is increasing, which is consistent to the line graph previous we plotted in section 1.
3. Limitations:
Due to the limitation of the data, there is no correlation between the type of environmental pollution and the change of temperature in that year. Hence in the visualisation, we can not break down the change of temperature into propotions which correspond to each of the pollution types.