DeepFeatures
About
Understanding Earth system dynamics and their response to climate change and human activity requires innovative approaches to analyse complex and multivariate remote sensing data. However, the current trend is towards large models that require a lot of memory and computational power to be trained. The DeepFeatures project addresses this challenge by developing a foundation model approach to create Feature Data Cubes, which capture the underlying ecosystem dynamics as a low dimensional representation in latent space. These reduced representations enable the use of simpler, resource-efficient downstream models, which are easier to train and require minimal computational resources.
The project builds on the rationale that each spectral index (SI) reflects a specific aspect of ecosystem behaviour. Despite the development of over two hundred spectral indices, current studies often narrow their focus to individual SIs, overlooking the broader context of land surface processes represented by not considered SIs. The DeepFeatures project addresses this challenge by adopting a spatio-temporal multivariate approach. The SIs are derived from Sentinel-2 observations to generate a SI Data Cube. AI dimension reduction methods are applied to reduce the SI dimensionality and extract a latent space to create the Feature Data Cubes.
To demonstrate the potential of the Feature Data Cubes, the project focuses on inference for a variety of science cases including:
- modelling gross primary production
- analysing tree mortality and greening trends
- biodiversity monitoring for conservation
- comparing phenological features using satellite and crowd-sourced data
- studying the ecological impacts of open-pit lignite mining.
DeepFeatures emphasises the deployment of transparent and reproducible workflows, from generating Sentinel-2 derived SI Data Cubes to creating Feature Data Cubes. It aims to have an accessible, extensible, and modifiable framework for diverse applications, fostering broad community engagement and enabling open exploration of Earth system dynamics.
The project is funded by the European Space Agency (ESA) as part of the AI4SCIENCE activity.
Data
- Sentinel-2 L2A
- Awesome Spectral Indices
- ERA5 Land
- ESA Climate Change Initiative (CCI) land cover classification gridded map
- Copernicus Digital Elevation Model
Team
- Karin Mora (Leipzig University, Germany)
- Julia Peters (Leipzig University, Germany)
- Martin Reinhardt (Leipzig University, Germany)
- Konstantin Ntokas (Brockmann Consult GmbH, Germany)
- Gunnar Brandt (Brockmann Consult GmbH, Germany)
- Teja Kattenborn (University of Freiburg, Germany)
- Guido Kraemer (Leipzig University, Germany)
- David Montero (Leipzig University, Germany)
- Clemens Mosig (Leipzig University, Germany)
- Daria Svidzinska (Leipzig University, Germany)
- Miguel Mahecha (Leipzig University, Germany)