GTAP Models

Overview
Current Model
Previous Versions
Dynamic Model
Other Models
RunGTAP

Research

Energy
Land Use and Cover
Baseline
Migration
Poverty

GTAP Research: Land Use and Land Cover

GTAP 9 Land Use and Land Cover Data Base

The GTAP 9 Land Use and Land Cover Data Base builds global land cover and land use databases for base years: 2004, 2007 and 2011. Unlike previous versions, the data is directly constructed from publicly available geospatial maps (circa 2000/01 at 5-minute grid resolution). These are then aggregated for each AEZ-region and updated to each base year using national level output price, land use and land cover information from FAOSTAT (2016).

Crop land use and land cover, pasture land cover as well as potential vegetation maps are based on the 5-minute grid resolution of the geospatial information used in constructing the original GTAP 6 LULC. This consists of agricultural land cover from Ramankutty et al (2008), land use data for 172 crops from Monfreda et al (2008) and potential vegetation from Ramankutty and Foley (1999) which are all available for download at www.Earthstat.org.

The administrative maps and agro-ecological zone maps (AEZs) in GTAP 9 LULC is different from other GTAP LULC versions. Administrative map is now based on Global Administrative Areas v2.8 (2016). New agro-ecological zones maps (AEZs) are computed from length of growing period and thermal zone information from the Global Agro-ecological Zones (GAEZ v3.0) by IIASA/FAO (2012). Gridded livestock information from FAO (2007) is also used to share-out land rents for ruminant livestock types.

The dataset is now available as part of the GTAP 9 Satellite Data and Utilities Package.

GTAP Research Memorandum No. 30, "Development of GTAP version 9 Land Use and Land Cover database for years 2004, 2007 and 2011", presents a detailed explanation of the methodologies used in developing the dataset as well as supplementary mapping files.

GTAP 8 Land Use and Land Cover Data Base
The GTAP 8 Land Use and Land Cover Data Base builds on global land cover and land use data bases, as well as global forestry data. In keeping with the multi-year release of version 8, the land cover and land use database is updated to two base years: 2004 and 2007.

The land cover data, available by land type, agro ecological zone (AEZ), and country for years 2004 and 2007 has been generated using the Global Cropland and Pasture Data by Navin Ramankutty from McGill University, Global Maps of Urban Extent from Satellite Data from the Center for Sustainability and the Global Environment at the University of Wisconsin-Madison and Global Maps of Potential Vegetation by Navin Ramankutty and Jonathan Foley.

Procedures in updating the land cover data
  1. Aggregation and normalization of agriculture and built-up land cover - Within each 0.5x0.5 degree grid cell the percentage of land cover for croplands, pasturelands and built-up lands were added. Grid-cells wherein the combined agriculture and built-up land cover exceeded 100% were normalized to ensure that the sum of land cover for these cells do not exceed 100%.


  2. Derivation of the "residual" land cover and its allocation to other land cover types - The "residual" land cover was calculated by deducting the normalized land cover of agricultural and built-up lands from grid-cells with uniform value of 100%. The "residual" land cover was then allocated to savanna and grasslands, forests, shrublands and other lands using the Global Maps of Potential Vegetation


  3. Conversion of the fractional land cover data into actual area of land cover - To derive the grid-cell level areas for the different land cover types, a global map of the surface area of the earth was used.


  4. Aggregation of land cover areas by AEZ and by country - The grid-cell level areas for different land cover types were aggregated by overlaying the land cover maps with a map of global AEZs and countries.


  5. Calculating accessible forest area by AEZ and by country - To derive the accessible forest area by AEZ and by country, the aggregated forest area data was weighted by the share of accessible forests in total forest area by AEZ and by country. These shares were taken from GTAP 6 land cover database.

Procedures in updating the Crop Production and Area
The new version of the GTAP Land Use and Land Cover Data Base has been developed using the Food and Agriculture Organization of the United Nations (FAO) 2004 and 2007 data on production, harvested area and price, available by country and 175 FAO crop categories. The data can be downloaded from the FAOSTAT website.

Development of the AEZ Land Rental Data Base
The further processing of the FAO data involves introduction of AEZs. The AEZ shares for production and harvested area are not available for years 2004 or 2007. Accordingly, these are computed from the GTAP 6 Land Use Data and are assumed to be unchanging over these periods. In other words, if maize production rose in the US over this period, it is assumed to rise in all AEZ regions by the same proportion.

For special cases when we observe non-zero production (harvested area) in FAO data base but zero AEZ shares in the GTAP version 6 (2001) land use data, we apply continental AEZ shares. There is no need for additional treatment of the land cover data, since it is already available by AEZ.

The dataset is now available as part of the GTAP 8 Satellite Data and Utilities Package.

GTAP Research Memorandum No. 23, "Development of a GTAP 8 Land Use and Land Cover Data Base for Years 2004 and 2007", presents a detailed explanation of the methodologies used in developing the dataset.


Applications





GTAP 7 Land Use Data Base
The GTAP version 7 land use data base builds on global land cover and land use data bases, as well as global forestry data.

The land cover data, available by land type, agro ecological zone (AEZ), and country for the year 2004 has been generated using the Global Cropland and Pasture Data by Navin Ramankutty from McGill University.

Procedures in updating the land cover data
  1. Conversion of the fractional land cover data into actual area of land cover - To derive the grid-cell level areas of croplands and pastures, a global map of the surface area of the earth was used.

  2. Aggregation of cropland and pasture land areas by AEZ and by country - The grid-cell level areas of croplands and pastures were aggregated by overlaying the crop and pasture land cover maps with a map of global AEZs and countries.

  3. Reconciling the changes in agricultural land cover with the land cover data in version 6 - Firstly, total lands across countries and across AEZs were preserved between the two versions. Secondly, changes in agricultural lands (croplands and pasturelands) were assumed to primarily affect forest lands.
Procedures in updating the Crop Production and Area
The new version of the GTAP land use data base has been developed using the Food and Agriculture Organization of the United Nations (FAO) 2004 data on production, harvested area and price, available by country and 175 FAO crop categories. The data can be downloaded from the FAOSTAT website (http://faostat.fao.org/).

Development of the AEZ Land Rental Data Base
The further processing of the FAO data involves introduction of AEZs. The AEZ shares for production and harvested area are not available for 2004. Accordingly, these are computed from the GTAP version 6 land use data and assumed to be unchanging over the 2001-2004 period. In other words, if maize production rose in the US over this period, it is assumed to rise in all regions by the same proportion.

For special cases when we observe non-zero production (harvested area) in FAO data base but zero AEZ shares in the GTAP version 6 (2001) land use data, we apply continental AEZ shares. There is no need for additional treatment of the land cover data, since it is already available by AEZ.

To compute the forest land area we use the GTAP version 6 forest land cover (accessible plus inaccessible forest land area) to share out the GTAP version 7 forest land across tree species, vintage, and management types. Thus we assume, as with crop production, that the regional composition of national forests does not change over this three year period. We then separate the accessible forest land from the total by applying the shares of accessible forest land from the version 6 data base.

GTAP Research Memorandum No. 19, "Development of the GTAP Version 7 Land Use Data Base", presents a detailed explanation of the methodologies used in developing the dataset. Also accompanying the memorandum is a land splitting utility which is designed to split the land rents across AEZs for any data base.