Global Climate Change: is it valid? — A look through both perspectives using EDA
For this module assignment, I’ve decided to revisit the global climate change dataset from Kaggle in order to investigate how both perspectives debate if climate change is valid or not through an EDA process. Prior to starting this analysis, I did some preliminary research for the major leading factors individuals use to argue that climate change is not factual. The most prominent evidence that the anti-climate change community uses is how the earth exhibits natural cyclical patterns of heating and cooling, and that some geological locations do not experience a large impact in temperature changes across time. The other representations of good-faith arguments for climate change is that due to CO2 emissions and human pollution there is a noticeable increase in global temperature and atmospheric change at an alarming rate. To determine which side is correct, or which perspective hold the most evidence based truthful claims, I will begin this EDA by examining the anti-climate change perspective to investigate some common claims.
We will start be uploading all of our imports before we do anything else.
Next up, we will load up and look through the different dataset that we will be using for this analysis. The data we will be using comes from Berkeley Earth, who released a number of datasets relating to climate change on Kaggle. I chose to use the city, country, and global land temperature data because these three datasets will be enough for us use EDA techniques to find conclusions that are made by each side of the argument for global climate change. After this, we will load up each one of the datasets that we will be using for this assignment.
To determine if there has been a significant difference in average land temperature over time, I will be using a simple .groupby() function to curate the top 20 hottest cities, as well as the countries. To visualize a change over time, I was only able to repeat the same process but change the “index.year” to equal 2010 so that I can get a 30 year difference from the same datasets.
You can see that there is only a+1.04 increase over a 30 year duration, making the overall impact minimal in terms of the geolocations that experience regular extreme heat. The hottest location did not change from Umm Durman in Sudan over the time duration, showing that global climate change has not made an extreme impact for the hottest geolocations on earth.
The next popular argument for anti-climate change community revolves around the earth naturally exhibiting patterns of global cooling and heating. Short term weather patterns are subject to fluctuate, which this data visualization helps illustrate this point by plotting the average city temperature for the United States from 1844 to 2000. As you can see from this plot, there are natural variance in weather patterns for small scale city geological locations. Small variance in the increase or decrease of temperatures is not enough for some to believe that global climate change is valid.
Biases: This perspective for global climate change validity challenges modern science by using segmented land temperature data from extreme geological locations which do not experience the most impact of global warming. Also, most cyclical patterns of intense temperature changes from global cooling and global warming operate on a much larger timescale rather than just between the past 50+ years that these datasets offer. There can also be interesting moral matrices in play for this perspective of thought, especially if an individual’s financial stability is reliant on the consumption of fossil fuels or if other businesses have high CO2 emission rates in order to operate. There can easily be confirmation bias that comes into play for both perspectives on global warming’s validity, but in order to understand the situation more clearly, we must investigate the perspective of climate change believers.
To begin our next perspective, we will look at the overall land temperature data to see if there has been any noticeable change over time for six random continents.
You can see from this chart that there are small variation across time, although some geological locations are impacted more than others. The United States has a noticeable increase from past 2000 onward. Other continents such as Australia and Greenland also see an increase in overall land temperature change, although it is not as noticeable as the U.S. This chart illustrates a definite increase in land temperature, but the timescale and number of continent variation is not sufficient enough to draw definitive conclusions.
Our the next step in our EDA process, we will draw from overall land temperature data instead of segmenting the data into top cities, countries, or continents. To do this, we will use a simple plot chart that curates the data to average global temperature changes between 1750 and 2010 to allow for a larger chronological timeline than previous illustrations.
From the data visualization above, you can see the temperature increase much clearer compared to the prior charts shown. You can see that there is a pretty strong variation across early years, but then there becomes a steady increase around late 1900’s and early 2000’s in which temperature begins to increase exponentially. This type of chart could be used to easily indicate that there is an abnormal global temperature increase throughout recent years, and could be validated even further by investigating CO2 emission data to determine what the leading factors are for human pollution and the overall impacts society has to global climate change. Lastly, I was curious to see what geological locations are impacted the most from climate change, so I used a plt. function to plot the top countries that experienced highest temperature differences between 1750 and 2010.
Canada, Russia, and Kazakhstan are the highest impacted countries from the effects of global climate change. I was surprised that the United States was ranked seventh, although it does raise questions on why exactly these countries are ranked highest over all other countries. My initial guess was that it would be from China, or the U.S. due to our mass market and consumption patterns.
Biases: Some potential biases for these good-faith representations for the global climate change argument could be that the evidence for global warming are so recent that there is not enough chronological data to make conclusive arguments yet. Although, the exponential rates of temperature changes and human pollution patterns does reasonably draw for considerable societal and international concern. I believe that this perspective is less likely to experience prominent moral matrices, but is highly susceptible to confirmation bias.
To conclude, my opinions on this subject still stand in support that global climate change is valid, although by doing this EDA assignment, I feel like I understand counter arguments more clearly. It is easy to skew data and have confirmation bias when researching, and the type of online engagement individual have I feel can heavily impact heir cognitive perspective on the matter. There are significant limitation based upon the datasets that I used, as most of them only indicate land temperature data from a limited chronological timeline, and does not contain many other exterior evidence to solidify any conclusive hypotheses. My prior assumptions were challenged by finding that extreme temperature locations such as Sudan and Saudi Arabia are relatively unaffected, which makes sense if some from these locations would not believe that climate change arguments are valid. It did not change my opinion, but it made me more understanding for how geological location can play a massive impact in an individuals perspective.
