Simulation of Global Warming in the Continental United States using Agent-Based Modeling

1
Simulation of Global Warming in the Continental
United States using Agent-Based Modeling

• By Marika Lohmus

2
Introduction

• Global Warming is a commonly accepted phenomenon
• The increasing global population has shown to
  increase the climate change
• More people more greenhouse gas output
• What about the effect of climate change on people?

3
Background

• No models involving population v. global warming
• School (Geosystems) STELLA models available to
  represent the global energy budget and change
• Similar research done in the Maryland Virtual
  High School using just System Dynamics and STELLA
• Various models using population in NetLogo
  (Damaceanu, An agent-based computational study
  of wealth distribution using NetLogo)
• No national temperature map data available

4
Abstract

• I will be using NetLogo for the project
• Representation of population changes (numerical
  and spatial where they move as the temperatures
  and sea levels increase)
• Visual Representation, Graphical Representation
• Able to change variables (such as CO2 output)
• System Dynamics model

5
Methodology

• Map of the United States
• Two variables
• Temperature
• Altitude
• Two agents
• Patches (land)
• Population (one agent per 1000)

Map using elevation Data
6
Interpolation and Averages

• For temperature mapping
• weatherbase.com for average yearly temperatures
  of major cities in the United States
• Interpolation linear relationship between a
  starting point and an endpoint (python)
• Not very accurate when applied to temperature
• Random mid-point temperatures mapped
• Averages if the upper and lower temperature was
  known, averaging was used to map out the
  mid-temperatures. (python)

7
1st Quarter Progress

• Firs quarter concentrated on just mapping out the
  temperature in the United States
• Much faster results when using python for
  interpolation

Current temperature map
8
Next Steps

• After temperature map
• Change CO2 levels
• See changing elevation and temperature values
• Include a graph and variable modifiers
• Equations from STELLA lab about global energy

• After temperature change
• Add population
• Death rates and birth rates
• Note the change of location of people
• Collect results

9
Population Agents

• Control Variables
• How high temperatures water levels they can
  withstand
• Other Variables
• Income according to 2000 Census demographic info
• Changing Variables
• Death rate

10
Problems

• The only issue is time.
• Temperature mapping is taking longer than
  expected
• NetLogo is very unforgiving about data errors
• If number of numbers in temperature data differs
  from the number of patches, it does not show
  where the error occurs.
• Only an extensive study of data will fix the
  problem
• If this takes even longer...
• Demographic data will not be included
• May concentrate project on just the environment

11
Possible Results

• Nearly impossible to predict it concerns the
  future
• Humans will move to northern/central United
  States
• Population will decrease
• Greenhouse gas input will decrease
• Temperatures / sea levels decrease
• Populations will increase and decrease in an
  oscillating fashion
• Perhaps an equilibrium will be reached

12
Possible Results Continued

• It is important to note that
• It is impossible to predict the future
• This project only concentrates on what will
  happen if nothing is done about the greenhouse
  gas output per person.
• The U.S. Temperatures mapped are yearly averages
• The population agents are at a ration of one
  agent per every 1000 people