Explore plant biodiversity at large scale and high resolution
GeoPl@nNet aims to make plant biodiversity easily accessible and understandable for everyone through high-resolution interactive maps and reports
Generate biodiversity report
GeoPl@ntNet is able to generate biodiversity report at a global “country” scale or at a “local” scale.
Global scale report
Get immediate biodiversity report on a whole country.
Local scale report
Draw shapes on the map to get report tailored to your areas of interest.
Explore species distribution predictions
Explore the predictions maps of species at a 50m resolution across Europe and compare them with maps of real occurences.
Dactylis Glomerata L.
France
Urtica Dioica L.
Germany
Explore habitats distribution predictions
GeoPl@ntNet lets you explore maps of EUNIS habitats predicted across Europe. The predictions are made at EUNIS level 3 habitats and can also be explored aggregated at both level 1 & 2.
Explore biodiversity indicators
GeoPl@ntNet lets you explore maps of biodiversity indicators across Europe.
Science behind GeoPl@ntNet
GeoPl@ntNet turns recent advances in deep learning and citizen-science into high-resolution, actionable biodiversity maps: it uses Deep-SDM convolutional models trained on remote-sensing, climate and occurrence data to predict species assemblages and derive biodiversity indicators, and a species→habitat transformer (Pl@ntBERT) to produce EUNIS habitat maps.
The approach draws on foundational research showing that convolutional neural networks greatly improve species distribution modelling by capturing environmental context, and on the Deep-SDM / Malpolon frameworks developed for continent-scale, 50×50 m mapping.
GeoPl@ntNet also stands on the Pl@ntNet family of citizen-science and computer-vision studies (the Pl@ntNet deep-learning system and the Pl@ntNet-300K image dataset), which provide the large, annotated observation corpus and identification models that underpin species-level inference.
About
GeoPl@nNet is developed and maintained by the Pl@ntNet team as part of a consortium agreement between CIRAD, INRAE, INRIA and IRD.
GeoPl@ntNet receives funding from the European Union Horizon Europe Research and Innovation Programme (ID No 101060693, 101060639). Views and opinions expressed are those of the author(s) only and do not necessarily reflect those of the European Union or the European Research Executive Agency (REA). Neither the EU nor REA can be held responsible for them.
