CIFOR Forest Day Event Bali December 8, 2007

Transcription

1 CIFOR Forest Day Event Bali December 8, 2007 Decision tools to identify where to focus interventions to reduce emissions from deforestation Sandra Brown Winrock International 1

2 Projections of deforestation at regional scales Have demonstrated the use of spatial analyses, using remote sensing data, key proxy drivers, and GEOMOD, to map threat for deforestation at regional scales Combined with historical rates of deforestation has been used to project deforestation over the medium to short term (applied to eight regions) Several lessons learned 2

3 Many variables affect spatial pattern of forest and non-forest lands Type of spatial data / Variable Infrastructure Biophysical Biological Human Disturbance Socionomic/Demographic Examples Distance to centers of government/commerce Distance to towns and communities Distance to roads (primary, secondary, all) Political district Elevation Slope Aspect Watershed Precipitation Temperature Distance to major rivers/navigable water Presence of wetlands Soil Vegetation type Ecozone Distance to logging camps Distance to forest edge Distance to previously cultivated areas Distance to previously deforested land Land tenure Density of land-based economic sector of population % marginalized population 3

4 No one set of proxy drivers explains patterns of deforestation Driver Sets Giving Best 'Goodness of Fit' Noel Kempff Guaraquecaba Rio Bravo Chiapas Elevation Slope Aspect Roads* # Rivers* Navig.Water* Political District Watershed Deforested Land * Communities* Centers of Commerce* Forest Edge * Soils Main Roads* WWF Ecozones Logging Camps* Ag. Pop. Density 0.21 Marginalization 0.20 Land Tenure LULC Distribution

5 General pattern of deforestation affects importance of different drivers Frontier pattern Forest non-forest mosaic Nonforest forest 5

6 Where should a country focus interventions to reduce emissions from deforestation? Spatial modeling useful tool to plan where to focus efforts to reduce deforestation to maximize return on investments 6

7 Steps used to simulate changes in land use/land cover using GEOMOD. Step 1 simulates the existing landscape and validates it with a reference land use map Step 2 simulates future land uses based on projected rates of deforestation and the Potential Land Use Change map (PLUC). Step 3 creates a carbon emission map over time based on a carbon map and simulated land use maps 7

8 East Kalimantan, Indonesia as a case study S. Petrova, F. Stolle (WRI), and S. Brown, 2007 Time 1 map Time 2 map 8

9 Examples of proxy drivers of land-use change 9

10 Main proxy drivers of deforestation in East Kalimantan Tested 102 combinations of drivers to explain deforestation in the region Kappa for Driver combinations used to create suitability maps location Accessibility driver maps Topography statistic Cities Deforested area Sawmills Rivers Roads Elevation x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x Accessibility is modeled as distance from.. 10

11 Threat map of future deforestation Combines suitability for change map with projected rates of deforestation over 10 year period generates a potential land use change map for period Rescaled based on equal interval of full range (1-210) into three classes 11

12 Combine with map of carbon stocks in above- and below- ground biomass pools Gibbs, H. and S. Brown Geographical Distribution of Biomass Carbon in Tropical Southeast Asian Forests: A Database. ORNL/CDIAC. 12

13 Identifies areas with potential high pay-off if well protected 13

14 Can overlay with other goals as needed to maximize multi-benefits Critical watershed Range of endangered species 14

15 Thank You! For more information see: Or contact me: 15