The VMERGE Data Management Team has made a first attempt at modelling Hporse density at 1km over Europe and to a lesser extent North Africa.
An ensemble of two models a Boosted Regression Trees (BRT) model and a Random Forest (RF) model was used to model horse density numbers at roughly 1km (0.00833333 degrees). The input horse numbers were taken from the Gridded livestock of the World database produced by the Food and Agricultural Organisation of the united Nations (Contact William.firstname.lastname@example.org if you would like to know more about the orifginal input dataset). The data was modelled/downscalled using VECMAP and the usual MODIS/WorldClim data-suite as used previously in modelling deer and boar.
There are two versions of the model. The raw ensemble model, and a "corrected" model for which the model densities have been forced to match the original input data at administration level.
Please note that this is a Beta product and first attempt so any feedback on this data from the project partners would be very much appreciated. Please not that for this reason the dataset is currently available only to project partners.