Gridded population databases: The demand side view

1
Gridded population databases The demand side
view

• Major usages
• Both descriptive and analytic
• Main areas
• Environment Biodiversity
• Climate
• Hazards Emergency management
• Land use change
• Agriculture
• Health
• Urban studies
• Economic development

2
The demand side view (continued)

• Minor usages
• Ready-to-go denominator
• Example used in Environmental Sustainability
  Indicator Project to adjust So2 emissions by
  populated land area rather than total land area
  (World Economic Forum, 2000)
• Proxy variable
• Example for climatic center of population
  distribution in a study of labor economics (Hall
  and Jones, 1998)

3
Better description about the distribution of
human population

• Cohen and Small
• World population generally localized along
  low-lying coasts, rivers
• Population peak at 2300 m altitude Mexican
  Plateau
• Data used GPW

4

Better description about the distribution of
human population (continued)

• UNDP WRI
• Population distribution (red) by aridity zone
  (browns)
• Shows highest density populations living in
  predominantly semi-arid or dry sub-humid climates
• Data GPW

5
Ecosystem stressHuman modification of coastal
areas

• WRI
• Close-up of SE Asia
• Global estimates nearly all areas with 100 km of
  coast are modified by human activity
• 20 highly altered by conversion to agriculture
  or urbanization
• Data Night-time lights (OLS)

6
Ecosystem stress (2)Water Availability by River
Basin

• WRI
• Estimate 2.3 billion people are living in
  conditions of water stress or water scarcity
• Allows for estimates by river basin not by
  national aggregates
• Many more than had been previously estimated
• Combines pop data with river basin model run-off
  data
• Data GPW

7
Health (1)Pop growth and the extinction of the
tsetse fly

• Project human and tsetse fly population
  distribution to 2040
• Human population growth causes loss of fly
  habitat
• Model human-fly interactions based on
  species-specific behavior
• Estimate that by 2040 the fly population will
  decline throughout Africa but an area as large as
  Europe will remain infested
• Pop growth affects subspecies differently

• Robin Reid et al., International Livestock
  Research Institute, Nairobi
• Combine human pop data with fly pop data
• previous efforts failed because national level
  data do not match that of fly habitat
• Data GPW

8
Health (2)Studies of Malaria

• Snow et al.
• Estimate morbidity and mortality in Africa
• 1 million deaths in 1995 estimated due to malaria
• 200 million clinical events
• Combine gridded population density with
  national-level data on age structure and malarial
  data (endemicity, hospital records)
• Data GPW

• Gallup and Sachs
• Deterministic analysis of malaria and economic
  growth
• Population density within 100 km a coastline--to
  proxy for access to transportation--is dependent
  variable
• Despite the strong correlation between poverty
  and malaria, and the strong impacts of malaria on
  the economy, the causal mechanism are unclear
• Data GPW

9
MigrationStudies of displacement of persons

• Dobson and colleagues
• Shows short term (1-2 years) population movement
  in Kosovo
• All systematic record keeping suspended
• Combines heuristic model (for pop data) with
  media accounts of wartime movement
• Data Landscan

10
Whats missing from known studies?

• Examples from
• Climate
• Land use change
• Urbanization
• Agriculture
• Hazards
• Economic development

11
Scale issues Small vs. large areas

• It seems that topics--and even disciplines--tend
  to have scale preferences
• Health studies tend to be local, at best
  continental
• Urban studies tend to be local or regional
• Climate studies tend to be global
• To what extent is this constraint data-driven
  rather than theory-driven?
• Can we create a collection of best-available data
  sets that can always be aggregated for
  coarser-scale analysis?
• To what extent are methods or inputs
  scale-specific?

12
Where have we come in 6 years?

• Lots of use
• UNEP/GRID-Environment Canada data base has had
  over 75,000 data transfers since 1996
• Dozens of published papers and books have used
  gridded population data
• almost all note that the study is improved or
  innovate in part because of the recent
  availability of population data on a grid
• Inputs are getting better
• Input data--both for population and
  administrative boundaries--continue to improve
• Satellite data are becoming more useful for
  integration with population data
• Methods are more sophisticated
• Applications are becoming more apparent
• Need for (and value of) interdisciplinary studies
  is
• increasingly recognized