I am concerned about the sustainability of our planet, and at times wonder what the climate and nature will look like in many (or not so many) years ahead. This winter has been a lot warmer than usual, and I’ve noticed climate change, in general, in the last 10 years of my life. I remembered seeing an ad once in the subway mentioning that places with more trees had cooler temperatures, and quite frankly, I hadn’t ever considered the impact that trees have on our climate. Given the concern on climate change and global warming, I wanted to focus on forest areas and climate change for this lab report and have the opportunity to discover patterns or trends. I started researching deforestation information and ended up finding a data set of the forest area per country beginning in 1990 through 2020. This data set has 9 years of annual forest area data for each country within that time frame of the years between 1990 and 2020. I additionally found global temperature anomalies data, with the expectation of finding patterns between temperature anomalies and forest area changes.
Prior to finding the data set I used, I browsed the internet and various sources. Through my sleuthing, I was inspired by visualizations I encountered, particularly from the Our World in Data website (https://ourworldindata.org). The inspirational visualizations spanned from charts to maps to tables, that helped me ideate on visualizations before I started the process.
Methods & Process
I first found a dataset through Kaggle that originated from the World Bank which included countries, their respective metadata and forest area in the last 30 years. I had to tidy the data via OpenRefine, since the years and forest area were appended as columns. After tidying the data, I loaded the source into Tableau. I began playing around with visualizations and was not finding the declines in the forest area that I had expected to see. I met with our professor, James Adams, during class time and he brought up some points and details I had overlooked. For example, the forest area for South America was less than Oceania, which doesn’t make sense, since South America is home to the Amazon, the largest rainforest in the world. This was a lesson I learned since I realized I had trusted a source despite having my doubts about the data I was seeing. Professor James was able to help me find a dataset that was more trustworthy, and similar to the one I had previously, from the Food and Agriculture Organization of the United Nations. James also provided feedback on the data type that was defaulted in Tableau (i.e. Year was pulled in as a “String” instead of “Date”), which was affecting how the chart represented the information. I later changed the data type for all my sources directly in Tableau and instantly saw the chart represented more accurately.
After getting the new dataset, I had to add specific metadata in order to visualize the data by certain categories, for example by continent. The new dataset had each country’s respective M49 code – a standard code used by the UN – only, along with its forest area. James helped find a source via the UN website that provided the list of countries, each with its respective M49 code and ISO code. I then used Google Spreadsheets to do a VLOOKUP formula to pull in the ISO Code via the M49 code into my existing dataset. But that wasn’t enough because I also had to pull in the continent information via the ISO code from another data table I found via Our World in Data. I therefore did another VLOOKUP formula, this time to pull in each country’s continent via the ISO code. I finally had the complete data table that I needed on the forest areas! There are nine yearly data points for this new dataset for each country. The dates start in 1990 and end in 2020.
As mentioned earlier, I myself have seen a huge change in temperature in the last, approximately, 10 years of living in the greater New York City area, which is why I was curious about adding the global temperature together with forest area and seeing whether there was any correlation. I had already found a dataset on the annual land and ocean temperature anomalies for the last 30 years via the National Centers of Environmental Information. I chose to use temperature anomalies since they are indicative of the changes in temperature based on a reference value, specifically the long term temperature average. This provides a gauge of climate change. According to the NCEI, “Anomalies more accurately describe climate variability over larger areas than absolute temperatures do, and they give a frame of reference that allows more meaningful comparisons between locations and more accurate calculations of temperature trends” (Global Surface Temperature Anomalies | National Centers for Environmental Information (NCEI), n.d.). Positive anomalies indicate a higher temperature from the long term average, while negative anomalies would indicate the cooler temperature in reference to the long term average.
Having all my data, I was now ready to begin using Tableau to chart the new data and information.
Summary of tools used:
– Google Spreadsheet
Summary of data sources used (with relevant columns):
– Food and Agriculture Organization of the United Nations (country, M49, year, forest area)
– United Nations Statistics Division (country, M49, ISO)
– Our World in Data (country, ISO, continent)
– National Centers of Environmental Information (year, global temperature anomalies)
Findings and Analysis
From my analysis, the most clear finding is that there are two continents where forest areas have considerably decreased since 1990: Africa and South America. Interestingly, Europe’s and Asia’s forest area increased since the beginning of the data’s time frame, but it was not as much as the aforementioned decrease in the other continents (see Figure 1). South America is home to the Amazon, the largest rainforest on the planet, which is significantly responsible for the health of our planet. According to a National Geographic article and video, one of the most important elements of the Amazon is that it contributes to temperature regulation and cooling of the planet (Haywood, 2022). The Amazon also absorbs one-fourth of the world’s CO2 emissions. As a result, the earth’s climate is impacted as more deforestation occurs in the Amazon. With continued deforestation, the rainforest will change to become arid and essentially cease to benefit our planet (Haywood, 2022).
I graphed the two data sources, forest area and temperature anomalies, through the “blend” affordance in Tableau. I used markers for both forest area and temperature data, where the shape’s size indicated the values of temperature anomalies and the color represented the forest area (Figure 2). Through this, I found the revelation I was expecting. I found that as the forest areas decreased, global temperature anomalies were higher, meaning the temperature was warmer in reference to the long term average. I separately plotted the temperature anomalies and noticed a pattern of constantly higher anomalies over the years, as recent years had a significant increase in temperature anomalies, again meaning the temperature is warmer in reference to the past. The anomalies have been increasing since 1990, and considerably jumped in the last 5+ years.
I ended up creating two dashboards with the several charts and a table that I had created showcasing the drastic changes in forest area and the climate anomalies. I shared the dashboard with my lab partner, Yeatasmin. Yeatasmin provided positive feedback on the highlight table, mentioning it was very clear, noting the drastic drop in forest areas for the South American and African continents via the stepped colors. I received helpful feedback regarding the two charts I had with similar data but minor differences in the visualization: one chart was a simple line chart, the other was a trend chart with square markers for each data value. Yeatasmin mentioned that the chart with markers was more transparent since that chart showed all the years in between that did not have actual data values (there’s a gap for example between 1990 and 2000). Yeatasmin suggested I explore treemaps to further showcase the changes in forest area from the start of the data, 1990, to the end value in 2020.
I took these suggestions and amended the dashboard. I removed the line chart with all continents. I added a different line chart that showed the bottom and top 2 continents that decreased and increased most, respectively. Filtering out the other continents eliminates visual noise and we can see better the net impact of forest area.
I also added two tree maps for 1990 and 2020, which gives another layer of visualization to the forest areas drops in Africa and South America, since the forest area changes are not only denoted in color but also in the size of the rectangles. I additionally thought it might be interesting to use a tree map for temperature anomalies, therefore I created one and added it to my dashboard. This tree map addition really outlines how much higher the temperature anomaly values are in recent years, which indicates that the temperature has been getting warmer from the reference point (historical average).
Overall, I think that there was success in the visual representation of the state of our forest areas and their impact on climate change. Since I am not very proficient with Tableau yet, I had some trouble blending the data at first, and had to iterate or recreate my visualizations several times. Though it was sometimes frustrating, I do appreciate the process because it allowed me to learn the nuances of the tool, and provided me with lessons learned with each “failure” or challenge. One task that I wanted to use Tableau for was to make data calculations on the fly (i.e. percent change) and then visualize those. I didn’t find a way to do this, but it would have possibly helped deepen the analysis. In general, I think the different visuals successfully told a story, which was my primary goal for this data visualization lab. I also learned a lot about the state of our forest areas globally, and the climate change values. Factually quantifying the information, aside from my anecdotal experiences, was eye-opening and really shook me up inside. This creates higher awareness about my own behaviors and actions that might impact our planet, ways I can adjust to waste less, and inspires me to share this information with others.
Haywood, A. (2022, December 12). Amazon Deforestation and Climate Change. National Geographic Society. Retrieved February 18, 2023, from https://education.nationalgeographic.org/resource/amazon-deforestation-and-climate-change/
Global Surface Temperature Anomalies | National Centers for Environmental Information (NCEI). (n.d.). National Centers for Environmental Information. Retrieved February 19, 2023, from https://www.ncei.noaa.gov/access/monitoring/global-temperature-anomalies
NOAA National Centers for Environmental information, Climate at a Glance: Global Time Series, published February 2023, retrieved on February 18, 2023 from https://www.ncei.noaa.gov/access/monitoring/climate-at-a-glance/global/time-series
Food and Agriculture Organization of the United Nations. Forest area (1990-2020 / 1000 ha). [Data Set]. https://api.data.apps.fao.org/api/v2/bigquery?sql_url=https://data.apps.fao.org/catalog/dataset/8dec11d0-2b46-481c-8b79-e5021d40cf0a/resource/790121b8-148f-4858-ab6e-a5ed3089adbb/download/fra-fa-query.sql
UNSD — Methodology. (n.d.). UNSD — Methodology. Retrieved February 18, 2023, from https://unstats.un.org/unsd/methodology/m49/
Continents according to Our World in Data. (2015). Our World in Data. Retrieved February 18, 2023, from https://ourworldindata.org/grapher/continents-according-to-our-world-in-data
Resources. Public.tableau.com. (n.d.). Retrieved February 18, 2023, from https://public.tableau.com/app/resources/learn