Prototypes and Feedbacks

Discussion

This set consistsof 2 different choropleth maps, which shows average temperature in different countries, over the years 1880 and 2013. For a single prototype, the color encoding is based on the average temperature variations. This set of visualizations helps to understand the climate change occurred between the years 1880 and 2013. As I am analyzing the earth's surface temperature changes, this visualization provides a clear picture of how this change has occurred across the world over a century. The planned interactivity will help understand more about the details of this gradual change.

Interactivity: There will be two associated interactivities for this set. First one is a slider which can be used to slide through different years ranging from 1880 to 2013. While sliding, maps will be updated with the corresponding year's data. Currently I have created the visualizations that will appear on the initial and final points of slider. The second interactivity will be hovering over each country, which will give a description about the country name and its average surface temperature.

Data and Processing

The data set GlobalLandTemperaturesByCountry.csv was used for this set of prototypes. It had 577,464 rows of data. The data processing was done using Tableau. The initial data had temperature recordings of different countries over different months, of years from 1743 to 2013. The data is missing for Antartica. Also data was missing for some countries beyond year 1880. So I chose to start the analysis from the year 1880. Using Tableau filtering and rollup, the average temperatures were calculated for each country for the years 1880 and 2013.

Feedbacks Received

The feedback for this visualization was given by Professor Sophie Engle. Main points from feedback are :

• Use of a slider won't show the temperature variation accurately as the color change won't be significant. So instead of a slider, calculate the temperature difference of each country between the start year and final year. This variation can be used for color encoding in choropleth.
  Final versions improvements: I have implemented this approach in my choropleth visualization.
• Adding an additional popup line graph to show the details of the temperature variation for each country, can be useful for detailed analysis
  Final versions improvements: I implemented this approach
• Avoid using diverging colors to reperesent temperature, as the 0 point cannot be defined as an exact value for each country.
  Final versions improvements: Changed diverging color gra to linear color gradient
• Modify the image size to avoid stretching in large screens
  Final versions improvements: I added max width tags for the images

Discussion

This set consists of 2 choropleth maps, which shows the average temperature of different states of US over the years 1890 and 2013. I chose US as the country to focus further on the analysis. The color encoding of each prototype is based on the average temperature of each state. This set will help compare, the temperatures in different states of US. Also the difference in color, over different years will show how the temperature has changed in US, over the past decade.

Interactivity: The planned interactivity is similar to the first set of prototypes. There will be a slider to show the temperature variations over the years 1890 to 2013. Currently I have created the visualizations that will appear on the initial and final points of slider. Also a mouse hover will show the details for a particular state.

Data and Processing

The raw data set GlobalLandTemperaturesByState.csv was used for creating this set of prototypes. The data was processed using Tableau. The data set had 645,677 rows, which had temperature recordings of different states in US, for years 1743 to 2013. But data was missing for most of the states beyond the year 1890. So I chose to start the analysis from the year 1890. The average temperature for each state, for 1890 and 2013 was calculated by rollup and filtering the data in Tableau.

Feedbacks Received

The feedback for this visualization was given by Professor Sophie Engle. Main points from feedback are :

• Since the first visualization already includes the data of United States, this set of visualization is not providing any additional information.
  Final versions improvements: I dropped the idea of implementing this visualization.

Discussion

This is a heat map showing the Earth's land temperature variation in degree celsius for the years 1850-2015 over different months. Heat map was chosen for the analysis as it encodes time series data efficiently. This is an important visualization for the analysis of climate change. The temperature here is encoded using color. The light colors show low temperatures and dark colors show the high temperatures. From the analysis of colors, we can see that there is a gradual increase in the color density. This shows that there is a gradual increase in Earth's surface temperature over the past years, which is the evidence of climate change.

Interactivity: The planned interactivity is, when mouse click on a row , the row is highlighted for the user, to focus on the particular month. Also on the side panel, it will show the temperatures in the years 1850 and 2015, and the difference in temperature.

Data and Processing

The raw data set GlobalTemperatures.csv was used for creating this set of prototypes. The data was processed using Tableau. The original data set had 3200 rows and 9 columns for the years 1750 to 2015 for different months, different days. But data was missing for some of the years and for some months. Since the average is used to represent a month's data, I filtered out the years with missing data, otherwise it can be misleading. Data was filtered by year, from 1850-2015, in 20 year intervals. The aggregation of data over a month was done by calculating average temperature for a month.

Feedbacks Received

The feedbacks for this visualization was given through peer reviews . Main points from feedback are :

• The temperature can be shown in fahrenheit instead of showing in degree celsius as fahrenheit is the common temperature scale used in US.
  Final versions improvements: I implemented this approach.
• The color gradient is indicating that the weather is getting colder over time instead of getting hotter. A better color gradient can be chosen to reduce the lie factor.
  Final versions improvements: I used orange color-linear gradient scale for the heat map.
• Make the color scale description better, saying it is showing the average temperature in degree celsius or fahrenheit.
  Final versions improvements: Added clear titles and legends.
• Can include maximum data that is available, not just the data in specific intervals to improve data density
  Final versions improvements: I included the data from 1755 instead of 1880.