Evolve AI Institute

Data Visualization Templates and Instructions

Lesson 6: AI in Climate Science and Prediction

Lesson 6: AI in Climate Science and Prediction

Table of Contents

  1. Introduction to Data Visualization
  2. Choosing the Right Graph Type
  3. Step-by-Step Instructions by Tool
  4. Best Practices for Climate Data
  5. Common Mistakes to Avoid
  6. Templates and Examples

Introduction to Data Visualization

What is Data Visualization?

Data visualization transforms numbers into pictures that make patterns, trends, and insights immediately visible. In climate science, effective visualization can communicate complex information to diverse audiences—from scientists to policymakers to the general public.

Why It Matters

AI's Role

AI helps create visualizations by:

Choosing the Right Graph Type

Decision Tree

What are you trying to show?

Change over time?
  └─> **Line Graph** or **Area Chart**

Comparison between categories?
  └─> **Bar Chart** or **Column Chart**

Relationship between two variables?
  └─> **Scatter Plot**

Parts of a whole?
  └─> **Pie Chart** (use sparingly!)

Distribution of data?
  └─> **Histogram** or **Box Plot**

Geographic patterns?
  └─> **Map** or **Heat Map**

Graph Type Guide for Climate Data

1. Line Graph

Best for: Showing trends over time

Climate Examples:

When to Use:

Strengths:

Limitations:

Example Data:

YearGlobal Temperature Anomaly (°C)
1880-0.16
1900-0.08
1920-0.27
1940+0.13
1960+0.03
1980+0.26
2000+0.40
2020+1.02
2023+1.15

2. Bar/Column Chart

Best for: Comparing distinct categories or discrete time periods

Climate Examples:

When to Use:

Strengths:

Limitations:

Example Data:

CountryAnnual CO2 Emissions (Million Tons)
China10,065
USA5,416
India2,654
Russia1,711
Japan1,162

3. Scatter Plot

Best for: Showing relationship between two variables

Climate Examples:

When to Use:

Strengths:

Limitations:

4. Area Chart

Best for: Showing cumulative values over time or proportion of total

Climate Examples:

When to Use:

Strengths:

Limitations:

5. Heat Map

Best for: Showing variations across geographic areas or time periods

Climate Examples:

When to Use:

Strengths:

Limitations:

Step-by-Step Instructions by Tool

Option 1: Google Sheets (Recommended for Beginners)

Creating a Line Graph

Step 1: Enter Your Data

  1. Open Google Sheets
  2. Enter data in two columns:
  3. Column A: X-axis values (Years, Dates, Time)
  4. Column B: Y-axis values (Temperature, CO2, etc.)

Step 2: Select Data

  1. Click and drag to select both columns including headers
  2. Make sure you've included column headers (important for labels!)

Step 3: Insert Chart

  1. Click Insert → Chart from menu
  2. Google Sheets will suggest a chart type
  3. If it doesn't choose line graph, select it from Chart Editor panel

Step 4: Customize Your Chart

In the Chart Editor panel (appears on right side):

Setup Tab:

Customize Tab:

Step 5: Make It Professional

Step 6: Download

  1. Click three dots in upper right of chart
  2. Download → PNG (best for presentations)
  3. Or Download → PDF (best for printing)

Option 2: Microsoft Excel

Creating a Bar Chart

Step 1: Prepare Data

Category    | Value
------------|-------
China       | 10065
USA         | 5416
India       | 2654
Russia      | 1711
Japan       | 1162

Step 2: Select Data and Insert Chart

  1. Highlight data including headers
  2. Click Insert tab
  3. Choose Column Chart or Bar Chart icon
  4. Select 2-D Column or 2-D Bar (simpler is better)

Step 3: Customize

  1. Click chart to activate Chart Design tab
  2. Add Chart Element:
  3. Chart Title: Click to edit
  4. Axis Titles: Add titles for both axes
  5. Data Labels: (Optional) Show values on bars
  6. Format Options:
  7. Right-click bars → Format Data Series
  8. Adjust Gap Width (20-30% for readability)
  9. Change colors: Click bar → Format Data SeriesFill

Step 4: Polish

Option 3: Online Tools (Free and Easy)

Using RAWGraphs (rawgraphs.io)

Best for: Creating unique, publication-quality visualizations

Step 1: Prepare CSV Data

Step 2: Upload to RAWGraphs

  1. Go to rawgraphs.io
  2. Click "Load your data"
  3. Paste or upload CSV

Step 3: Choose Visualization

Step 4: Map Your Data

Step 5: Customize

Step 6: Export

Using Datawrapper (datawrapper.de)

Best for: Quick, professional-looking charts and maps

Step 1: Create Account (Free)

Step 2: Upload Data

Step 3: Check & Describe

Step 4: Visualize

Step 5: Annotate

Step 6: Publish & Export

Best Practices for Climate Data

The 5 C's of Effective Visualization

1. Clear

Good: Obvious what data represents at first glance

Bad: Requires explanation to understand

Tips:

2. Correct

Good: Accurate data, appropriate graph type, proper scale

Bad: Misleading scales, wrong graph for data type, errors

Tips:

3. Complete

Good: Includes all context needed for understanding

Bad: Missing units, source, time period, or key information

Essential Elements Checklist:

4. Concise

Good: Communicates efficiently without unnecessary elements

Bad: Cluttered with decorative elements that don't add meaning

Tips:

5. Color-Smart

Good: Colors enhance meaning and are accessible to all

Bad: Random colors, or color choices that exclude color-blind viewers

Color Guidelines:

Good Color Palettes for Climate Data:

Common Mistakes to Avoid

Mistake #1: Truncated Y-Axis (Exaggerating Change)

Problem: Starting Y-axis above zero makes small changes look dramatic

Example of Misleading:

Temperature (°F)
68°                           ●
67°              ●
66°    ●
     2020      2021      2022

Looks like huge increase!

Corrected:

Temperature (°F)
80°
60°    ●      ●           ●
40°
20°
0°
     2020   2021      2022

Shows actual modest change in context

When truncated axis IS okay:

Mistake #2: Pie Chart Overload

Problem: Pie charts difficult to compare; bad for more than 3-4 categories

Better Alternative: Use bar chart for precise comparisons

When pie charts work:

Mistake #3: 3D Effects

Problem: 3D charts distort data, make values hard to read, look unprofessional

Solution: Always use 2D charts

Why: Science communication values accuracy over flashy effects

Mistake #4: Too Many Variables

Problem: Trying to show everything on one chart creates confusion

Example: Line graph with 15 different countries = spaghetti mess

Solution:

Mistake #5: Missing Context

Problem: Data without source, time period, or units is meaningless

Must Include:

Climate Data Visualization Templates

Template 1: Global Temperature Change Over Time

Data Structure:

Year | Temperature Anomaly (°C)
-----|-------------------------
1880 | -0.16
1890 | -0.26
1900 | -0.08
... (every decade)
2020 | +1.02

Recommended Visualization: Line graph

Title: "Global Average Temperature Anomaly (1880-2023)"

Key Features to Show:

Color Scheme:

Caption: "Global average temperatures have risen 1.1°C since 1880, with most warming occurring after 1980. Data from NASA Goddard Institute for Space Studies."

Template 2: CO2 Concentrations (Keeling Curve)

Data Structure:

Year | CO2 Concentration (ppm)
-----|------------------------
1960 | 316.91
1970 | 325.68
1980 | 338.75
1990 | 354.39
2000 | 369.55
2010 | 389.90
2020 | 414.24

Recommended Visualization: Line graph with seasonal variation

Title: "Atmospheric CO2 Concentration at Mauna Loa Observatory"

Key Features:

Color Scheme:

Template 3: Arctic Sea Ice Decline

Data Structure:

Year | September Sea Ice Extent (Million km²)
-----|---------------------------------------
1980 | 7.8
1990 | 6.2
2000 | 6.3
2010 | 4.9
2020 | 3.9

Recommended Visualization: Area chart or line graph with range

Title: "Arctic Sea Ice Minimum Extent (September)"

Key Features:

Color Scheme:

Template 4: Renewable Energy Growth

Data Structure:

Year | Wind (GW) | Solar (GW) | Fossil (GW)
-----|-----------|------------|------------
2010 | 200       | 40         | 4800
2015 | 435       | 228        | 5200
2020 | 743       | 714        | 5500

Recommended Visualization: Stacked area chart

Title: "Global Energy Capacity by Source (2010-2020)"

Key Features:

Color Scheme:

Pro Tips from Data Visualization Experts

Edward Tufte's Principles

"Above all else show the data." - Edward Tufte, data visualization pioneer

  1. Maximize data-ink ratio: Every mark on your chart should represent data
  2. Avoid chartjunk: Remove decorative elements
  3. Use small multiples: Show many small charts instead of one complex chart
  4. Respect the truth: Don't distort data for effect

Climate Communication Best Practices

From IPCC Guidelines:

From NASA Visualization Guidelines:

Assessment Checklist

Use this to evaluate your visualization before submitting:

Data Accuracy (5 points)

Technical Elements (10 points)

Visual Clarity (10 points)

Communication (5 points)

Additional Resources

Interactive Tutorials:

Books:

Climate Data Sources:

© 2025 Evolve AI Institute LLC. May be reproduced for educational use.

Lesson 6: AI in Climate Science and Prediction