New agro-ecological zones and their potential to affect land use projections

Transcription

1 New agro-ecological zones and their potential to affect land use projections Alan V. Di Vittorio Lawrence Berkeley National Laboratory GTSP/GCAM Meeting 2 October 2013 CLIMATE & CARBON SCIENCES PROGRAM

2 2 Global distributions of Paddy Rice Production

3 3 Different boundaries give different local estimates C 0.5 degree (~50 km) per side C C 0.25 degree (~25 km) per side C 1 degree (~100 km) per side Temperature maximum, Jan. 1, 2003 Cell size is 2.5 minutes (~5 km)

4 4 IAMs have different regions/land units Unquantified spatial uncertainty confounds intermodel comparison and ensemble analysis Model Regions Land units for use projection IMAGE (RCP 2.6) 26 half-degree grid MiniCAM (RCP 4.5) 14 GCAM: 151 land units AIM (RCP 6.0) MESSAGE (RCP 8.5) 24 half-degree grid 11 half-degree grid

5 5 Different land use/cover representations in ESMs obscure land use change effects on regional climate Uncertainty chain: IAM land use spatial uncertainty Land use/cover translation ESM land cover Temperature effect of RCP 8.5 land use change for (Brovkin et al. 2013)

6 6 In the context of the integrated Earth System Model (iesm) How do we make robust projections of land use change in the context of projected climate change? How do spatial boundaries influence projected land use?

7 7 Overview What are agro-ecological zones and why do they matter? Current versus projected GCAM land units Effects of projected land units on GCAM AgLU crop and land rent inputs

8 8 Agro-Ecological Zones (AEZs) are bio-climatically defined

9 9 Original agro-ecological zones Tropical Temperate Boreal

10 10 Land use distribution assumes uniform vegetation productivity within zones Tropical Temperate Boreal

11 11 Current land units become heterogeneous

12 12 Current land units become heterogeneous

13 13 Data required to create new AgLU crop and land rent inputs Spatially explicit data VMAP0 countries (246) AEZ countries (160) SAGE data: crop yield, area cropland pasture land area potential vegetation HYDE3.1 data: urban land area AEZ boundaries Tabular data GTAP countries (226, 87) FAO countries (241) GTAP (SAGE) crops GTAP use sector GTAP land rent FAO crops FAO crop production FAO producer prices FAO crop yield, area for recalibration

14 14 Workflow to create new AgLU crop and land rent inputs Data Aggregate original land rents by use sector to 87 GTAP countries Identify land cells Optional: recalibrate to different FAO data year Disaggregate crop land rents to 18 AEZs based on production and price Calculate crop production and harvested area per 18 AEZs X 226 GTAP countries Disaggregate forest land rents to 18 AEZs based on original land rents and forest area

15 15 Global distributions of Maize, by country Production (t) Harvested area (ha)

16 16 Distribution differences for Maize, by country Production difference (%) Harvested area difference (%)

17 17 Distribution differences for Oil Palm Fruit, by country Production difference (%) Harvested area difference (%)

18 18 Distribution differences for Paddy Rice, by country Production difference (%) Harvested area difference (%)

19 19 Distribution differences for Wheat, by country Production difference (%) Harvested area difference (%)

20 20 Global distributions of Maize, by land unit Production (t) Harvested area (ha)

21 21 Global distributions of Oil Palm Fruit, by land unit Production (t) Harvested area (ha)

22 22 Global distributions of Paddy Rice, by land unit Production (t) Harvested area (ha)

23 23 Global distributions of Wheat, by land unit Production (t) Harvested area (ha)

24 24 Each crop is uniquely affected by new land units Wheat Rice Oil Palm Maize

25 25 Summary AEZ-based land units do not consistently meet homogeneity assumption for land use projection New AgLU input system performs better than GTAP with respect to FAO data Distributions of crop production and harvested area are different between the original and new land units Unique crop responses to land units, technology, or climate will affect land use projections Feedbacks: climate, impact, and land use

26 25 Summary AEZ-based land units do not consistently meet homogeneity assumption for land use projection New software performs better than GTAP with respect to FAO data Distributions of crop production and harvested area are different between the original and new land units Unique crop responses to land units, technology, or climate will affect land use projections Feedbacks: climate, impact, and land use

27 25 Summary AEZ-based land units do not consistently meet homogeneity assumption for land use projection New AgLU input system performs better than GTAP with respect to FAO data Global distributions of crop production and harvested area are different between the original and new land units Unique crop responses to land units, technology, or climate will affect land use projections Feedbacks: climate, impact, and land use

28 25 Summary AEZ-based land units do not consistently meet homogeneity assumption for land use projection New AgLU input system performs better than GTAP with respect to FAO data Distributions of crop production and harvested area are different between the original and new land units Unique crop responses to land units, technology, or climate will affect land use projections Feedbacks: climate, impact, and land use

29 25 Summary AEZ-based land units do not consistently meet homogeneity assumption for land use projection New AgLU input system performs better than GTAP with respect to FAO data Distributions of crop production and harvested area are different between the original and new land units Unique crop responses to land units, technology, or climate will affect land use projections Feedbacks: climate, impact, and land use

30 Questions? Tropical Temperate Boreal