Geo Viz Critiques

[rasagy]

Hi B.Des. & M.Des. students! Now that we’ve gone through different type of geo-visualizations (slides here!), let us find examples of these online and learn from them.

Each group will pick three geo-visualizations, and mention:

1. Topic of the viz
2. Geo viz type
3. Dataset and individual properties, along with the type of data (ordinal/nominal/quantitative/qualitative)
4. Visual encoding for each property
5. Critique (How effective is it?)

You can add one comment for each group, and share these info and the link here.

[RishabhMakes]

Group 9:

Visualisation 1

Topic: The Relation between the Schematic map of the Berlin Subway and the actual map of Berlin Subway Geo Viz Type: Lines on a Map Dataset and type: The data is the different tracks of the Subway system. The different tracks are a nominal data type Visual Encoding: Different routes of the system are plotted in different colours Critique: This is effective in showing how the schematic map transforms into the real one but doesn't provide a way to effectively compare them like an overlap.

Visualisation 2

Topic: Proportion of homes with a thatched roof Geo Viz Type: Choropleth Dataset and type: Proportion of thatched roof houses in every district (Quantitative), District name(Nominal) Visual Encoding: The proporton is represented using Saturation Critique: The visualisation has some strong points. The data is broken in intervals and encoded in saturation. This helps to reduce clutter and enables easier pattern recognition. To save from the loss of exactness in doing so it also provides the exact figure on hovering. This aids in direct comparison and analysis amongst districts.

Visualisation 3 (Legend)

Topic: Accumulation of Crime in Chicago Geo Viz Type: Animated 3D Histogram Dataset and type: Date and time of Crime(Temporal), Crime Type(Nominal) Visual Encoding: The type of crime is encoded using color, the number of instances are encoded in length Critique: The number of categories is a hinderance that hasn't been taken care of using ways like grouping or labelling. The 3D nature of the histograms make it virtually impossible to compare different heights of the bars owing to parallax effect and the color coding used.

[VarunVikash]

Group 4

Vishnu(M.Des), Varun, Pranjal

Visualization 1MQ~!CN*OTI4NzI2Mw.NzE4NTgxNg)

Topic True size of countries on the Mercator map Geo Viz Type Cartogram Dataset and type Area of countries - Quantitative Position of countries on a Mercator map - Qualitative Visual Encoding Hue - To differentiate between countries that are selected Area&Skew - Representation of a country on the map if it was at the location specified by the user Critique Ability to drag the countries around makes the understanding and basic visual comparison easier, but numbers are not very accessible as each country must first be selected via the search bar so as to be able to view the popup of information on hover. Opacity of the country block ensures that the all countries are visible at all times despite the user's actions. Availability of clear map feature and description are not evident since they are grouped with social media icons.

Visualization 2

Topic Predicted movement of animals due to climate change Geo Viz Type Flow Map Dataset and type Location of current and predicted habitats - Qualitative Number of animals moving out and moving in for each habitat - Quantitative Predicted path of movement - Qualitative Category of animal - Nominal Visual Encoding Hue - Animal category Dot - Group of animals with same start and end points Length of trace left by dot - Number of animals in that group Path taken by dot - Predicted path of the group Critique It is very difficult to identify the points where dots originate form and hence almost impossible to track the movement of a particular group and the lack of any form of marking other than country name makes it difficult to discern the areas of movement. The locations of the predicted habitats, however are relatively easier to locate due to the convergence of the moving dots. Since the data is available only for the the Americas, it would be better to make it evident or best, use a map containing only North and South America so as to avoid the viewer mistaking the unavailability of data for a technical error and wasting time on it. The use of the word 'migration' in the title without the mention of climate change in the title can stands for something entirely different than what the visualization is trying to convey.

Visualization 3

Topic What Powers the World?(Source of electricity) Geo Viz Type Dots on Map Dataset and type Electricity consumption of each area - Quantative Source of electricity for each area - Qualitative Some facts about energy production and consumption - Qualitative Visual Encoding Each dot - A specific unit of electricity consumed Density of dots - Amount of electricity consumed in an area Circle - Location of a country that holds an interesting fact Critique In its current form, since energy from all sources is represented by the same dot, it is not possible to compare energy coming from two different sources without having to use the toggle switches. A simple encoding of different sources to different hues would have solved this issue. The boundaries of countries in well-lit areas are not at all visible due to the border being of same colour as that of the dots, making it difficult to compare between countries.

[naveenshaji]

Group 8

Visualisation 1

Topic

Arms sales: USA vs USSR since 1950

Visualization Type

Flow

Data Type

Quantity of arms sold over a period. (Quantitative) Which country they are being sold to. (Qualitative)

Visual Encoding

Dots denote units in TIV. The rate of flow denotes the quantity sold over a certain time. Red dots are arms coming from the USSR and Russia(later). Blue dots are those coming from USA.

Critique

• Estimation of constantly moving dots is very difficult. A static flow diagram with variable widths would have been more effective in conveying the data.
• The bar graphs on the bottom left corner seem disconnected from the geo-visualization. Size indicators on the map itself may have been a better choice.

[smokeybot7]

Group 3: Angela, Richa, Nebin

1. Topic of the viz London’s Oyster Card tidal flow by Oliver O'Brien Geo viz type Dot map timelapse Dataset and type Quantitative data type The dots show touch-ins and touch-outs of Oyster cards at London’s tube and train stations. The area of the circles (not radii) are directly proportional to the absolute number of taps in each 10-minute interval. The data is from a weekday except Monday/Friday to avoid the weekend edge effect. Visual encoding for each property Red dots: touch-ins Green dots: touch-outs Critique Color code could be swapped as green is usually associated with ‘entry’ and red with ‘exit’. Not really clear why opacity changes.

2. Topic of the viz Smellmap of Amsterdam by Kate McLean Geo viz type Isopleth Dataset and type Qualitative data type Dots mark the origins of the smells, concentric circles indicate their range and the warped contours allude to potential smell drift in the north- and south-westerley winds encountered on the days of the smellwalks. Visual encoding for each property The colors for each smell was inspired by the city’s visual landscape.

3. Topic of the viz Animated Tornado Tree Rings by John Nelson Geo viz type Circular Bins with tree rings Dataset and type Quantitative data type The individual rings in one tree ring indicates the intensity. Smallest ring is the weakest and largest ring is the most intense. The position of the rings corresponds to the geographical location on the map. Visual encoding for each property The colors of the rings are the different colors assigned to 12 months of a year. Critique Visually interesting but too many colors can get overwhelming to interpret. There are versions with lesser colors just for the 4 seasons.

[ArunJRK]

Group 7: Vishnu, Thuli, Arunjayaramakrishnan 1.US Health data

Viz Type: Choropleth Data-set type Quantitative data of patients death and different diseases, visualized in the form of choropleth Encoding method Divergent color pallet is used to encode the magnitude of the data.

2.Taxi rides per hour in a week Viz Type: Bin n Heat-map Data-set type Quantitative data of no of rides taken in New york city by hour in a week. Data is taken from gps and data points are average of 90millon data points Encoding method Divergent color pallet from yellow to red to violet.

3.Country-to-country net migration (2010-2015) Viz Type: Flow Map

Data-set type Quantitative data of the net immigration of all countries for 2010-2015.

Encoding method Colours of the circles are used to distinguish net loss or gain, and the area of any circle gives us an idea of the magnitude. Moving yellow dots denote people and their rate of flow denotes the number of people moving in or out of the particular country.

Critique

• Getting a sense of a quantity with moving dots is a bit taxing.
• The dots don’t seem to have any representational value, as some places the dots’ number and speed is arbit.

[r1ckrck]

Group 2: Arnesh, Varsha and Advait.

1. City Ways

Geo viz type Line map time-lapse

Dataset and type Quantitative data The thickness + intensity of the lines represent the density of the people either running/walking/cycling on the path. the map features a time-lapse through months of the year, hence showing change in the density across the year.

Critique A tooltip which shows the actual value of the density on a certain road would be helpful to interpret the map better. Right now, even thought the overall picture is clear, one cannot really decipher the actual values of density on a certain road.

2. Roboat

Geo viz type Bins

Dataset and type Nominal Data and Quantitative data The bins represent a certain unit area of the map. Clicking on one off the bins selects the bin as a starting point of a journey. Then hovering over other bins tells you the time taken to travel to that place using a certain mode of transport. the bin is highlighted using the fastest mode of transport that could be used to reach there from the starting point. The different hues of the boxes represent the different modes of transport.

Critique It could be communicated better that the colour of the box represents the fastest mode of transport.

3. Toilets in tehsils

Geo viz type Colopleth

Dataset and type Ordinal data A convergent colour scale did used and the hue is also changed along the scale for better identification.

Critique Only gives an overall idea, fails to convey detailed information.

[kapadiaparth]

Group 5: Parth, Jonathan & Aarcha

---

Visualization 1 Here’s every total solar eclipse happening in your lifetime. Is this year your best chance?

Geo Viz type: Connections/Lines on a map.

Dataset and type:
Nominal Data of the path of totality for the various eclipse encoded in form of connections between different points on Earth. The hue of the connection represents the time from now when the eclipse is supposed to occur. The darker one's are to occur in near future.

Critique: The viz is focused on major cities of the world. Its very much an overview. Another piece of information which is missing is that when during a day would it be occurring. Would be really nice if we could actually get a count of how many eclipse which a country might see during the entire span of the scope of the viz.

---

Visualization 2 LIGHTS ON | LIGHTS OUT A Global Look at Where Our Night Lights Have Turned On or Dropped Out

---

Visualization 3 Five years of drought

[enlinquental]

Group 6 : Enlin, Kuldeep & Manas

Visualization 1 The Food Capitals of Instagram

Geo Viz type : Dots on map.

Dataset and type : Place where the photo of the food item was hashtagged - Qualitative How much percentage of the food item was hashtagged at that place - Quantitative

Visual Encoding : Dot - place where the photo of the food item was hashtagged. Size of the dot - percentage of the food item hashtagged at that place. Saturation of the dot - higher value means good impact. Hue - the chosen food’s continent of origin.

Critique : The dataset of food items is limited and shows only popular ones. It does not have a timeline, which could show the trend of the food, when and where was the major influence occured in a period, ranking of the food at a particular time and how the hashtag counts were influenced by season, festivals etc.

Visualization 2 Here’s How America Uses Its Land

[Abhi98krishna]

Group 1 : Abhijith, Kiran, Shikha

Visualization 1: Wind Map

Geo Viz type : Line flow map

Dataset and type : Wind speed and direction of wind flow: both Quantitative.

Visual Encoding : Line - Shows the path of wind flow. Density of lines - Speed of wind

Critique : Gives a holistic understanding of the wind flow and wind speeds all over the United States. Zoom feature allows us to view wind speeds in particular locations. But there is a lack in information when it comes to understanding how and where the wind flows from or to. The intention of the map was to help understand the potential for wind energy capturing, which is well captured through this Geo-viz.

Visualization 2: Average commute times

Geo Viz type : On zooming in: Choropleth On zoom out: Dot map

Dataset and type : Average commute time: Quantitative

Visual Encoding : Diverging color palette indicating pace of traffic Choropleth area based on pin code area

Critique : One of the major problems with this visual representation was the diverging color palette. The palette uses dark purple for 'fast' and light yellow for 'slow'. We intuitively connote darker colors to be slow, heavy, deep, and lighter colors with fast and light. Hence a reversal of the use of color would add value to the vizualization. Another information which lacks clarity is the commute points. the commute times are given, but without any mention of whether it is within a single pin code, with the neighboring areas or of the people living in the area.

Visualization 3: Twenty years of India

Geo Viz type : Dot map, along with a line graph on the bottom showing the light variation with time.

Dataset and type : Amount of light emission : Quantitative

Visual Encoding : Opacity of color denoting intensity of light

Critique : The map gives a good overall idea of light emission in the country. It allows for comparison with previous years as well as a comparison between different places in India. Interesting insights come out of the comparison between developed and developing areas, and these can be made quickly, thanks to the apt encoding of opacity and density. The line chart, however does not convey any information clearly

[rasagy]

These are an inspiring set of projects, and thanks for writing down the discussions that were done in the class — going ahead and closing this ticket.