PROMISE Work Package 2100 Impact of LandUse Changes on Future Monsoon Climates

1
PROMISE Work Package 2100Impact of Land-Use
Changes on Future Monsoon Climates

• The development of realistic scenarios of land
  cover change in tropical Africa for the period
  1960 2050
• Dermot ORegan
• CEH Wallingford

2
Study Objectives
WP2100 To use global and/or regional control and
scenario integrations to address the impact of
land use changes (e.g. deforestation) on monsoon
climates. The sensitivity of the results to the
representation of hydrological and land-surface
processes will be investigated.
CEH Wallingford Scenarios for land cover change
from 1960-2050 will be generated for sub-Saharan
Africa based on simple assumptions about
population growth, land use practice, and current
rates of deforestation. Gridded changes in
generic vegetation classes (forest, cropland,
pasture etc.) will be converted to changes in
land surface parameters which can be applied
consistently in different GCMs. Comparison
between modelled responses will be used to aid in
the identification of common mechanisms and
model short-comings.
3
Presentation Outline
1. The Problem Data and Understanding -
highlight the problems of obtaining reliable
estimates of land cover and land cover change
essential for global change monitoring

• 2. Study Method
• - present the methodology and some results of
  this part of WP2100

4
Data and Understanding

• Land Cover Land Use
• Reliable, comprehensive and timely information on
  global land cover and land use is essential in
  the context of global change research
  (DeFries and Townsend 1994)
• Land Cover Land Use Change
• Understanding the dynamics of land-use and
  land-cover has increasingly been recognised as
  one of the key research imperatives in global
  environmental change research (Lambin 2001)

5
Figure 1 A comparison of estimates of extent of
ten cover types derived from three existing
ground-based datasets of global land
cover (from DeFries Townshend 1994)
6
Study Method

• collate data relating to present land cover and
  deforestation rates in the study region
• identify the area of each land cover type, by
  country, both in terms of actual hectarage and as
  a percentage of total (national) land area
• generate scenarios of past and future land cover
  in the region
• map land cover classes into GCM functional types
  tree, shrub, grass, bare soil - for use in
  different GCMs
• make data available for integration with
  meteorological and hydrological components of CEH
  work

7
Study Region Tropical Africa
West Africa (16) Senegal Gambia Guinea Bissau
Guinea Sierra Leone Liberia Ivory Coast
Ghana Togo Benin Nigeria Chad Niger Burkina
Faso Mali Mauritania
Central Africa (6) D.R.C. Congo Gabon Eq.
Guinea Cameroon C.A.R.
8
Global Land Cover Data Sources

• FAO
• FAOSTAT
• Global Forest Resources Assessment
• (FRA1990, FRA2000)
• World Forestry Inventory
• (1953, 1958, 1963)
• UNEP
• Environmental Data Report
• (1987 - biennial)
• World Bank
• World Development Report
• (1977 annual)

• World Resources Institute (WRI)
• World Resources
• (1984-biennial)
• Central Intelligence Agency (CIA)
• World Factbook
• (1962-2001)
• USGS
• Global Land Cover Characteristics Database (IGBP
  Land Cover Classification)

9
Varying Estimates of Land Cover
Problems with land cover and land cover change
estimation - definitions of land cover types -
reliable and consistent observation and
inventory - inaccessibility (ground or satellite
based methods) - remote sensing classification
10
FAO Global Forest Resources Assessment
2000(FRA2000)
1990 Land Cover Estimates
11
FAO Global Forest Resources Assessment
2000(FRA2000)
12
Scenarios of Land Cover Change

• Using the FRA 2000 baseline data, scenarios of
  land cover were generated for 1961, 2025 and 2050
  based on FRA 2000 land cover classification
• the change in forest extent was estimated for
  each country in the region (increase back to
  1960s, decrease into 21st century)
• For the 1961 scenario forest extent was altered
  according to land cover change estimates from the
  Environmental Data Report 1991/92 (UNEP) covering
  the period 1966 1988
• Future forest cover scenarios were generated
  based on two models, and applied to the years
  2025 and 2050. A BAU (business-as-usual) model
  gives forest cover estimates based on the
  scenario of a continuation of current rates of
  deforestation, as determined by FRA 2000
• A population growth model generated an
  alternative future forest cover scenario whereby
  annual deforestation rates (1990-2025 and
  2025-2050) are based on a relationship between
  population density and cumulative forest loss as
  published by Pahari and Murai (1999)

13
Scenarios of Land Cover Change
Forest Loss 1960 - 2050
14
Scenarios of Land Cover Change

• The extent of land cover within the other two
  classes was then increased, or decreased,
  accordingly. A partition of 40 other wooded
  land and 60 other land was applied, based on
  the proximate causes and specific agricultural
  practices causing deforestation in tropical
  Africa as assessed by Geist and Lambin (2001).

15
Scenarios of Land Cover Change
Land Cover Change 1960-2050
16
Mapping Land Cover Classes to GCM Functional
Types
Following expert guidance, FRA2000 land cover
classes were converted to functional types
tree, shrub, grass, bare soil - for application
in different GCMs
17
Conclusions

• Reliable, consistent, and timely data on the
  dynamics of land cover and land use are an
  essential component in global change research
• Currently, estimates of land cover vary greatly
  and the understanding of the drivers of land use
  change remains a key issue
• For this study scenarios for land cover change
  from 1960 -2050 were generated for a study region
  of tropical Africa based on simple assumptions
  about population growth, land use practice, and
  current rates of deforestation
• Will results show that climate and hydrological
  models are sensitive to the estimated changes in
  land cover and land use presented?