Frequently Asked Questions

  1. How do I get the data?
    The data is available for download from our Data Download page, where you will also find a description of the various datasets available. The LUH2 data is also available from the CMIP6 site.
    Note: The Land-Use Harmonization 2 (LUH2) datasets are made freely available for use by the scientific community, with attribution. Please use the following citation in any presentations or publications that result from, or include, the use of the LUH2 datasets:
    G. Hurtt, L. Chini, R. Sahajpal, S. Frolking, et al. “Harmonization of global land-use change and management for the period 850-2100”. Geoscientific Model Development (In prep).
    A DOI will be generated at the time of publication.

  2. What are land-use states?
    The land-use states are the fractions of each grid-cell occupied by various land-uses in a given year. In the original Land-Use Harmonization we used 5 land-use states: cropland, pasture, primary, secondary, and urban. These land-use states are also used in the second generation Land-Use Harmonization products, although we now subdivided several of these states into sub-categories too (e.g. 5 crop types within the cropland state, two pasture types within the pasture state, and both primary and secondary land subdivided into forested and non-forested lands).

  3. What are land-use transitions?
    The land-use transitions are the annual changes between land-use states. The land-use transitions give the fraction of each grid-cell that transitions one from land-use state to another in a given year. In addition to agricultural expansion and abandonment, land-use transitions include wood harvesting, shifting cultivation and crop rotations.

  4. What are primary and secondary land?
    In the Land-Use Harmonization we define primary land as natural vegetation (either forest or non-forest) that has never been impacted by human activities (e.g. agriculture or wood harvesting) since the beginning of our simulation (the initial year of the simulation depends on the dataset being used). Secondary land is also natural vegetation (either forest or non-forest) that is recovering from previous human disturbance (either wood harvesting or agricultural abandonment). Secondary land can range from very young vegetation recovering from a recent human disturbance, to vegetation that is very mature, recovering from a human disturbance many years ago, and which could be similar in biomass density to primary vegetation of the same type. In our model, there are no transitions to primary land; secondary land can never return to a “primary” state, even if it is very mature.

  5. How can I compute the forest area in a given year?
    For the original Land-Use Harmonization products, first you will need to download the forest/non-forest map (fnf_map.txt) and the grid-cell area map (cellarea_halfdeg.txt) in addition to the LUH data packages. The forest area in an individual grid-cell is given by: (gothr + gsecd)*fnf*cellarea (where gothr and gsecd are the fractions of the grid-cell occupied by primary and secondary land respectively). To compute the global forest area you would sum this quantity over all grid-cells globally.
    For the second generation of Land-Use Harmonization products, the primary and secondary land is already sub-divided into the forested and non-forested components. You will also need to use the grid-cell area which is part of the staticData_quarterdeg.nc file associated with the dataset. The forest area in an individual grid-cell for these products is then given by: (primf + secdf)*carea. To compute the global forest area you would sum this quantity over all grid-cells globally.

  6. What is difference between land-use and land-cover?
    Land cover describes the physical state of the land, whereas land use describes the human use of that land. We provide land-use data, which can be used in a broad range of models using different land-cover classification schemes. The use of the land can often imply something about the underlying land cover (e.g. "cropland" often has quite different characteristics from forest cover) but the relationship between land-use and land-cover can often be ambiguous (e.g. "pasture" can imply a land-cover of irrigated grassland, or a natural land-cover such as savanna). Our modeling process does not use a land-cover map or classification scheme but it does make assumptions about the underlying potential biomass density of the natural vegetation and the locations of potentially forested land, and these data layers are also available for download (secondary mean biomass density is included in all LUH datasets, while primary biomass density and the forest/non-forest map are included either under Supporting Files, for the original LUH products (see pot_biomass.txt and fnf_map.txt), or within the staticData_quarterdeg.nc file (for second generation LUH products). In addition, our second generation LUH products now sub-divide both primary and secondary land into forested and non-forested land to provide a simple, first-order land-cover classification.

  7. What are the secondary mean biomass and secondary mean age data layers?
    When primary land is harvested for wood, or agricultural land is abandoned, that land becomes secondary land of age 0, and with a mean biomass density of 0. Each subsequent year that secondary land grows and increases in age and biomass density (according to the underlying potential biomass density and regrowth rate), as well as changing in area as secondary land is added to, or removed from, each grid-cell. Our model tracks these changes and computes an estimate of the mean biomass density (gssmb in the original products, secmb in the second generation products) and mean age (ssma in the original products and secma in the second generation products) of the secondary land in each gridcell. The model uses these data layers as part of the algorithm to spatially allocate wood harvesting. These data layers are available as part of each Land-Use Harmonization dataset.

  8. What is the difference between "managed pasture" and "rangeland" in the new LUH products (LUH2)?
    In our new products, the total grazing land is sub-divided into "managed pasture" and "rangeland", based on the underlying input datasets from HYDE. This disaggregation of grazing land is based on an aridity index and population density index (dryer and less populated land is more likely to be classified as rangeland).

  9. Who should I contact with questions?
    For project questions and support please contact any of the primary researchers involved in the project: lchini@umd.edu, gchurtt@umd.edu, steve.frolking@unh.edu, and ritvik@umd.edu.

  10. Where can I get more information?
    In addition to contacting us, you can get more information from the README files associated with the data (LUH2_v1.0h Readme File, LUH2_v0.3 Readme File, LUHa.v1 Readme File, LUHa.v1_future.v1 Readme File, LUHa.v1_future.v1.1 Readme File), and from our publications on the Links page, especially the publication on this project that appeared in a special issue of Climatic Change: Hurtt et al. 2011. Also, don't forget to sign up for our mailing list, to receive future updates from us.