Academy of Finland  
Funding decision
Applicant / Contact person Vanhatalo, Jarno
Organisation University of Helsinki
Project title Multivariate Gaussian processes for hierarchical modelling of species distributions
Decision No. 317255
Decision date 18.06.2018
Funding period 01.09.2018 - 31.08.2022
Funding 450 732
Project description
Species distribution models describe the dependences between a species and their environment. They are a key part in ecologists' toolbox, for example, to study species habitat preferences, to manage conservation areas and natural resources and to predict effects of climate change. Currently, the state-of-the-art SDMs are joint species distribution models (JSDMs). This research proposal will be in the forefront of that development by developing semiparametric JSDMs. A widely accepted significant deficiency in (J)SDMs is that they are restricted to analyzing survey data only. This can significantly decrease their reliability in predictions concerning changing environment since surveys typically do not cover the whole covariate space of interest. Hence, we will develop methods to combine complementary sources of information in JSDMs to obtain better data coverage of the covariate space where predictions are made. Moreover, we will develop free software to be distributed to applied users