SUPR
Data-driven biodiversity analyses
Dnr:

NAISS 2024/5-18

Type:

NAISS Medium Compute

Principal Investigator:

Tobias Andermann

Affiliation:

Uppsala universitet

Start Date:

2024-02-01

End Date:

2025-02-01

Primary Classification:

10612: Biological Systematics

Allocation

Abstract

As humans we have a profound negative impact on the natural world surrounding us, leading to experts declaring the current biodiversity crisis. However, to date it is difficult to impossible to quantify how negative our impact on biodiversity really is. This ability to quantify biodiversity and the loss thereof is a big contemporary challenge for the biological research community. AI models have the potential to majorly aid in this task, as they allow estimating complex biodiversity dynamics by integrating different data types. In my research I am developing neural network models that can be used to model the capacity of a given site to harbor biodiversity. I am also working on Bayesian Neural Network models that can model vegetation-changes through time, directly quantifying uncertainty in the prediction.