DAVI Visualizing Geospatial Data

Lecture Notes: Geovisualization and Spatial Data Visualization

Announcements

• Project Presentations: Sign-up is open for project presentations in the TA sessions next week or the week after. Opportunity to showcase progress and receive feedback.
• Exam Sign-up Phase 2: If you have obligations during the exam period, email the professor to receive the sign-up link early.
• Office Hour Cancellation: No office hour today due to the professor teaching another lecture.
• Adjusted Schedule: Breaks will be shortened to 10 minutes to end the lecture 10 minutes early.

---

Recap Quiz on Temporal Visualization

Question 1: Horizon Graphs

• Purpose: Used to show multiple time series.
• Example: Stock prices over time or weather data for different cities.
• Visualization Technique: Compact display allowing comparison across many time series.

Question 2: Time Arrangements

• Not a Time Arrangement: Logarithmic time.

• Actual Time Arrangements:

  • Linear Time: Standard timeline from start to end.
  • Cyclic Time: Time represented in cycles, such as clock faces for daily patterns.
  • Branching Time: Represents diverging paths, like version control systems (e.g., Git branches).

Question 3: Triangular Model

• Identification: The chart shown is a triangular model.
• Purpose: Visualizes multiple time intervals as points instead of lines.
• Contrast with Other Charts:
  • Ternary Plot: Used for three variables that sum to a constant (e.g., 100%).
  • Population Pyramid: Shows age distribution across genders.
  • Tri-linear Plot: Another term for Piper diagrams in geochemistry.

Question 4: Mapping Time

• Not a Principal Way: Mapping time to pixel.
• Principal Ways of Mapping Time:
  1. Mapping Time to Space: Using small multiples to show different time points in separate spaces.
  2. Mapping Time to Time: Creating animations where display time represents data time.

Question 5: Spiral Plots

• Purpose: Used to identify periodicities in data.
• Example: Visualizing recurring patterns in events such as doctor visits or weather phenomena.
• Technique: Adjusting the spiral to match the period of interest (e.g., weekly cycles).

---

Geo-visualization

Introduction

• Origin of Visualization: Geo-visualization, or thematic cartography(专题制图), is considered the origin of visualization.
• Fundamentals: Many visualization principles, such as the use of marks (points, lines, areas), stem from cartography.
• Importance: Understanding geovisualization is crucial due to its widespread application and foundational concepts.

Challenges in Geo-visualization

Misrepresentation in Maps

• Example: The Mercator projection distorts areas, making regions near the poles appear larger.
• Consequence: Can lead to misconceptions about the size and importance of different regions.

Election Maps and the Modifiable Areal Unit Problem (MAUP)

• Issue: Traditional election maps may misrepresent data due to varying area sizes of regions.
• Example: A large area with a small population may appear more significant than it is.
• Solution: Use cartograms or other visualization techniques that adjust for population or electoral influence.

Trade-offs in Visualization

• Expressiveness vs. Effectiveness: Balancing accurate representation of geospatial context with the clarity of data presentation.
• Expressiveness of Data Context vs. Data Content: Deciding whether to prioritize geographical accuracy or data accuracy.

---

Map Projections

Purpose of Map Projections

• Challenge: Representing the 3D Earth on a 2D map introduces distortions.
• Types of Distortions: Area, shape (angles), and distance.

Common Map Projections

Mercator Projection

• Characteristics:
  • Preserves angles (conformal).
  • Distorts areas, especially near the poles.
  • angle（角度）指的是地球表面上两条相交线之间的夹角
  • 当地图投影被称为conformal（保角投影）时，意味着它在局部范围内保持了这些角度的准确性。
  • 
• Use Cases:
  • Nautical navigation: Straight lines on the map correspond to constant compass bearings.
  • Online maps (e.g., Google Maps) for local navigation.

Mollweide Projection

• Characteristics:
  • Preserves area (equal-area).
  • Distorts shapes, especially at the edges.
• Use Cases: Representing global distributions where area accuracy is important.

Azimuthal Equidistant Projection

• Characteristics:
  • Preserves distances from a central point.
  • Distorts areas and shapes elsewhere.
• Use Cases: Radio and seismic mapping from a specific location.

Evaluating Projections with Tissot’s Indicatrices

• Technique: Placing identical circles on the globe to visualize distortions when projected onto a map.
• Interpretation:
  • Size Changes: Indicates area distortion.
  • Shape Changes: Indicates angular distortion.

Quality Criteria for Map Projections

1. Angle Preservation (Conformal): Shapes of small areas are preserved.
2. Area Preservation (Equal-Area): Relative sizes of areas are maintained.
3. Distance Preservation (Equidistant): Distances from a central point are accurate.

• Note: It’s impossible to preserve all three simultaneously; projections must compromise based on use case.

---

Types of Maps in Data Visualization

Dot Maps - POINTS

• Usage: Plot individual events or occurrences at specific geographic locations.
• Example: Locations of disease outbreaks or crime incidents.

Heatmaps - POINTS

• Purpose: Visualize the density of points in an area to address overplotting in dot maps.
• Example: Population density or Wi-Fi signal strength in an area.

Flow Maps - LINES

• Usage: Show movement or connections between locations using lines or arrows.
• Example: Airline routes or migration patterns.

Choropleth Maps

• Characteristics:
  • Data is mapped onto predefined regions, such as political boundaries.
  • Uses color shading to represent data values.
• Appropriate Data: Data that is aggregated or inherently linked to regions (e.g., election results).

Isopleth Maps

• Usage: Represent continuous data fields by connecting points of equal value (isolines).
• Example: Weather maps showing temperature gradients.

Chorochromatic Maps

• Purpose: Display categorical data across regions.
• Example: Soil types, land use categories, or vegetation zones.

---

Multivariate Data on Maps

Using Glyphs

• Technique: Place complex symbols (glyphs) on the map to represent multiple variables.
• Example: Star glyphs showing literacy rates, income, and population in different regions.

Small Multiples

• Approach: Displaying a series of maps, each showing different variables or subsets of data.
• Example: Maps showing demographic changes over several decades.

Embedding Maps in Charts

• Method: Incorporate合并 geographical shapes into other chart types.
• Example: Using state shapes in a scatter plot to represent data points.

---

Adjusting the Map to the Data

Cartograms

• Definition: Maps where the sizes of regions are distorted to represent data values.

• Types of Cartograms:

  • Contiguous Cartograms: Maintain shared borders but distort shapes.
  • Non-Contiguous Cartograms: Regions are resized but do not maintain shared borders.
  • Dorling Cartograms: Use circles scaled to data values, placed approximately where regions are located.
  • Demers Cartograms: Similar to Dorling but use squares.

  

  

  

Creating Cartograms

• Algorithmic Approach:

  • Use a cost function with terms for area discrepancies, unused space, shape distortion, topology violations, and displacement.
  • Adjust weights in the cost function to prioritize different aspects.

  

Grid (Unit) Cartograms

• Characteristics:
  • Represent each region with a grid cell of equal size.
  • Useful for visualizing data where each unit (e.g., electoral votes) is equally significant.
• Example: Election maps where each grid cell represents one electoral vote.

Metro Maps

• Purpose: Simplify complex transit routes for easier navigation.
• Features:
  • Straight lines and evenly spaced stations.
  • Abstracted geography to emphasize connectivity over exact location.
• Design Variations:
  • Hexalinear and Octolinear: Use specific angles for route lines.
  • Curvilinear: Use curves instead of straight lines.

Map Schematization

• Objective: Simplify map shapes while retaining recognizability. 简化图形并且保持可以辨认的特性
• Technique: Reduce the number of points defining a region’s boundary, possibly replacing lines with arcs.
• Application: Creating clearer visualizations by removing unnecessary details.

---

Combining Time and Space

Space-Time Cube

• Concept: A 3D cube where the base represents geographical space, and the vertical axis represents time.
• Usage: Visualizing trajectories or movements over time in space.
• Example: Tracking the path of a vehicle over time.

Travel-Time Maps (Isochrone Maps)

• Purpose: Show areas reachable within specific time frames from a starting point.
• Features:
  • Isochrones: Lines connecting points of equal travel time.
  • Visualization: Colors or contours indicating travel time intervals.
• Example: Maps showing commute times to a city center.

---

Applying Spatial Visualization Beyond Geospace

Sports Visualization

• Example: Heatmaps on a basketball court showing player movements or shot locations.
• 

Game Maps

• Usage: Visualizing data in virtual spaces like game levels.
• Example: Dot maps showing player deaths in a Counter-Strike map.
• 

Other Applications

• Chess Boards: Mapping moves or strategies across the board.
• Indoor Spaces: Visualizing foot traffic or evacuation simulations within buildings.

---

Visualization Critique

Critique 1: Unit Cartogram of Beer Consumption

Issues Identified

• Color Scale Misuse:
  • Used a categorical color scale for sequential (ordinal) data.
  • Black representing high values is counterintuitive.
• Awkward Placement and Gaps:
  • Regions like Sweden not correctly placed.
  • Unexplained gaps for countries like Switzerland and Norway.
• Trade-off Consideration:
  • Sacrificed geospatial accuracy for the ability to place labels in small regions.
  • May hinder the ability to quickly identify regions based on geographic knowledge.

Recommendations

• Correct Color Scale:
  • Use a sequential color scale that intuitively represents increasing values.
• Improved Placement:
  • Arrange regions more accurately to their geographical location.
  • Label gaps or include placeholders for missing data.
• Expressiveness vs. Effectiveness:
  • Find a balance that maintains geographic context while effectively conveying data.

Critique 2: Line Chart of Republican Presidential Polling Data

Issues Identified

• Distorted Time Axis:
  • Only plotted dates when polls were taken, leading to uneven time intervals.
  • Misrepresents the duration of trends.
• Color Usage:
  • Not colorblind-safe; red and green used together.
  • Insufficient distinction between colors for different candidates.
• Missing Labels:
  • One of the candidate lines is not labeled, causing confusion.

Recommendations

• Time Axis Correction:
  • Use a continuous time axis with interpolation between polling dates.
• Colorblind-Friendly Palette:
  • Implement color schemes that are distinguishable for all viewers (e.g., ColorBrewer palettes).
• Labeling:
  • Ensure all data series are properly labeled.
  • Consider highlighting remaining candidates and de-emphasizing those who have dropped out.
• Alternative Visualization:
  • Explore using stacked area charts to represent voter percentages summing to 100%.
  • Use small multiples or separate charts for clarity if necessary.

---

Conclusion

• Importance of Geovisualization: Essential for accurately and effectively conveying spatial data.
• Careful Design Choices: Critical to consider map projections, color scales, and representation techniques to avoid misinterpretation.
• Balancing Trade-offs: Must weigh expressiveness, effectiveness, and efficiency in visualization design.
• Inclusivity in Design: Use color schemes and labels that are accessible to all audiences.

---

Additional Resources

• ColorBrewer: colorbrewer2.org - Tool for selecting colorblind-safe color schemes.
• Map Projection Animations: G.Projector - Software for exploring map projections.
• Visualization Principles: Tufte, E. R. The Visual Display of Quantitative Information - Foundational text on effective visualization.