4
Elaborate

Data Visualization

Learning Objectives

Why Visualization Matters

"The greatest value of a picture is when it forces us to notice what we never expected to see." - John Tukey

Chart Selection Guide

Data TypeQuestionRecommended Chart
Time seriesHow does concentration change over time?Line plot, area chart
DistributionWhat is the shape of the data?Histogram, box plot, violin plot
ComparisonHow do groups differ?Bar chart, grouped box plot
RelationshipAre two variables related?Scatter plot, bubble chart
CompositionWhat are the parts of a whole?Stacked bar, pie chart (if few categories)
SpatialHow does air quality vary by location?Choropleth map, heat map
Temporal patternAre there diurnal/seasonal cycles?Polar plot, calendar heat map

Principles of Effective Visualization

Do

  • Use clear, descriptive titles
  • Label axes with units
  • Start y-axis at zero (usually)
  • Use color meaningfully
  • Include data source
  • Keep it simple
  • Consider colorblind accessibility

Avoid

  • 3D effects (distort perception)
  • Excessive decoration ("chartjunk")
  • Truncated axes without indication
  • Too many categories/colors
  • Dual y-axes (can mislead)
  • Pie charts for >5 categories
  • Rainbow color scales

Air Quality-Specific Visualizations

Polar Plots

Display concentration as a function of wind direction and speed. Useful for identifying pollution source directions.

Pollution Roses

Show frequency of concentrations from different wind directions. Indicates which directions bring polluted air.

Calendar Heat Maps

Display daily values in calendar format, colored by concentration. Reveals day-of-week and seasonal patterns.

Diurnal Profiles

Average concentration by hour of day. Shows traffic rush hours, photochemical production (ozone), and other temporal patterns.

Color for Air Quality

AQI Color Scale

Good
0-50
Moderate
51-100
USG
101-150
Unhealthy
151-200
Very Unhealthy
201-300
Hazardous
301+

Sequential scales: For continuous data (light to dark for low to high)

Diverging scales: For data with meaningful midpoint (e.g., above/below standard)

Audience-Appropriate Design

Public/Media

  • Simple, familiar chart types
  • Clear takeaway message
  • Minimal technical jargon
  • Bold colors and labels

Policy Makers

  • Comparison to standards
  • Trend information
  • Geographic context
  • Uncertainty indication

Scientists

  • Technical detail acceptable
  • Statistical annotations
  • Multiple panels OK
  • Full methodology noted

Activity: Visualization Portfolio

Using one year of hourly PM2.5 data from a monitoring station, create:

  1. Time series plot: Show the full year with 7-day moving average and NAAQS annual standard line
  2. Distribution visualization: Histogram or box plot comparing weekend vs. weekday concentrations
  3. Calendar heat map: Daily averages colored by AQI category
  4. Diurnal profile: Average concentration by hour, with weekday/weekend comparison
  5. Summary graphic: Design a one-page infographic suitable for community presentation

Evaluation criteria: Appropriate chart type, clear labels, effective use of color, accurate representation, visual appeal.

Key Takeaway

Effective data visualization transforms air quality data into understanding. The best visualizations match chart type to data structure, follow design principles that enhance rather than obscure meaning, and are tailored to their audience. In an era of abundant data, the ability to create clear, compelling, and honest visualizations is an essential skill for communicating environmental information.

← Lesson 3: Machine Learning Lesson 5: Data Project →