Vegetation maps are useful for research and land-use and natural resource management, but often very expensive to produce for large areas. Horvath and colleagues from GEco and UiO Geosciences explore distribution modelling of vegetation types as an alternative in this new paper.
The authors conclude that the national distribution of some vegetation types can be captured better than other types; their predictions are better for rare vegetation types than common ones, and better for coastal vegetation types than inland ones.
By unlocking the potential of distribution modelling of vegetation types, researchers and natural resource managers alike can access vegetation maps for large areas that include ecological information not offered by remote-sensing methods. Further, distribution modelling also holds potential for large-scale mapping using international surveys, and for using the results to improve vegetation parameterisation in Earth System models.
Read the paper for more details. DOI: 10.1111/avsc.12451. Maps, results and code are available on GitHub and DRYAD (https://doi.org/10.5061/dryad.nk3b5k8)
