Data from "Spatial filtering strategies for mitigating sampling bias in species distribution models"

Results obtained and analysed in Lamboley & Fourcade, Spatial filtering strategies for mitigating sampling bias in species distribution models. Briefly, we used two virtual species with contrasting levels of specialisation to explore the impact of spatial filtering distances on the performance of ecological niche models. This investigation was conducted across a spectrum of modelling conditions, encompassing diverse types and degrees of bias, as well as varying sample sizes.Results reporting the overlap between modelled and true distributions:Unbiased_distribution.csv: results for the models trained from unbiased, i.e. randomly sampled, datasetsBiased_corrected_distribution.csv: results for the models trained from biased datasets, corrected with various spatial filtering distancesResults reporting the overlap between modelled and true response curves:Unbiased_response_curves.csv: results for the models trained from unbiased, i.e. randomly sampled, datasetsBiased_corrected_response_curves.csv: results for the models trained from biased datasets, corrected with various spatial filtering distances