Earth Science meets AI: Research within KI:STE

Cloud systems | Renewable Energy | Regional Climate - Dwaipayan Chatterjee

From an energy perspective, cloud systems play a significant role in modulating the solar radiation budget and the availability of solar radiation on the ground. They thus are of high interest for renewable energy applications. From a climate change perspective, they influence the distribution of heat and moisture around the planet. Significant uncertainty exists about their physical characteristics and missing links across various spatial and temporal scales. Satellites capture a field of clouds, and very often similar, looking individual clouds form neighbors of each other, giving rise to spatial patterns. But from a satellite and AI perspective, we may have many cloud system observations; still, they need to be labeled/segmented to train a conventional supervised neural network. Therefore, from an AI point of view, my research focuses on learning without labels and letting the neural network learn from scratch, understand the underlying representation of satellite observations, and focus on extracting meaningful features which can be used further for data-driven process understanding of regional climate and exploit its usage for solar energy power production.

Left: Classification by the self-supervised deep neural network at k = 7 for 128 x 128 configuration. Thirty random samples were selected over central Europe and visualized for each class. Each image is assigned a unique class through a colored frame. The bar charts in the lower part of a) and b) represent the number of images in relative percentage and cloud fraction in each regime. To better associate the class number with the cloud regime, the centroid image is shown as well. 

Right: Feature space with wet-dry conditions defined based on 80/20th percentile of integrated water vapor and w-500 wind speed from ERA-5. Grey defines dry conditions, blue is wet conditions, and light grey is the intermediate state, dry/wet intrusion as decided based on the 30 nearest neighbors. Average profiles of vertical wind speed W, and sea surface temperature (SST) for the reference states, intrusions, and their 30 nearest opposite neighbors.

Terrestrial Protection - Timo Stomberg

Monitoring the unique characteristics of protected areas, such as wilderness areas and national parks, is essential for effective management and conservation efforts. Of course, these characteristics can also occur in non-protected areas and it would be a valuable tool to produce maps highlighting areas with similar patterns.

Typically, explainable machine learning tools are used to explain the decision-making processes of models. But it can be also used to gain insights into satellite data, such as multispectral Sentinel-2 imagery. For this purpose, we build a specific convolutional neural network, with which we apply several attribution methods in a new way. This enables us to generate high-resolution maps that represent the complex patterns of protected areas in a comparable manner.

Model Error Correction Kaveh P. Yousefi

This research, conducted as part of the KISTE project, addresses the challenge of improving atmospheric data predictions in fully coupled models by merging model-based data with observations. The motivation behind this study stems from the uncertainties present in numerical weather prediction and climate models. From a broader perspective, the research aims to learn from all the mismatches (i.e., non-linear time-space error structure) between atmospheric model simulations and reference data. The challenge lies in defining a deep learning (DL) network capable of learning the model-reference mismatches (δ) for a set of variables from a model (m) and a set of variables from a reference data (r).

TSMP model (left bottom) incorporates various component models such as Parflow for subsurface, CLM for land surface, and COSMO and ICON as weather prediction models using the OASIS coupler. Weather prediction models often produce erroneous precipitation predictions due to factors like initial value and model errors. These errors negatively impact the water and energy balances in the land and subsurface, introduce feedback in fully coupled simulations, and generate inaccuracies in impact modeling. By merging atmospheric data, specifically precipitation, with observations, the accuracy of other simulation data within the system can be improved. Learning from model errors enables a more comprehensive understanding of the entire system and enhances the overall accuracy of predictions. The figure on the right demonstrates significant improvements achieved by correcting original model-based precipitation and surface pressure data using reanalysis data as the reference.

Landcover and Crop Classification - Ankit Patnala


Land cover and crop classification are essential for environmental monitoring, land use planning, and agriculture management. However, obtaining labeled training data is challenging. Self-supervised contrastive learning methods have emerged as a promising approach for learning representations from unlabeled data. These methods can be used to classify land cover and crops without requiring labeled data. We investigate the effectiveness of self-supervised contrastive learning methods for land cover and crop classification using Sentinel-2 and Plantescope data. All available channels in the data give a comprehensive view of the landscape and crops. Overall, this research aims to contribute to the development of more efficient and accurate land cover and crop classification methods by leveraging self-supervised contrastive learning methods and high-resolution remote sensing data. The results of this research have important implications for environmental monitoring, land use planning, and agriculture management.

– ChatGPT

Shallow Landslides - Ann-Kathrin Edrich


Shallow landslides are a common natural hazard in mountainous regions, and their prediction is crucial for mitigating risks to infrastructure and communities. We aim to create static and dynamic hazard maps for shallow landslides in Switzerland using only openly available data and a random forest machine learning algorithm. The research will leverage a range of openly available data sources, including digital elevation models, soil information, and climate data, to create a comprehensive dataset for training a random forest model. The model will be trained using historical landslide data from Switzerland and validated using cross-validation techniques. Static hazard maps will identify areas that are susceptible to landslides based on a range of static factors, such as topography and soil properties. Dynamic hazard maps will take into account additional factors such as rainfall patterns and soil moisture to identify areas that are at higher risk of landslides in real-time. To validate the accuracy of the hazard maps, the research will compare the predicted hazard zones with historical landslide events in Switzerland. The research will also investigate the sensitivity of the hazard maps to different factors, such as the spatial resolution of the input data and the number of trees in the random forest model. The resulting maps will provide valuable information for decision-makers and stakeholders involved in land use planning and disaster risk reduction.

– ChatGPT

Universality in Deep Learnning - research by Thomas Seidler


The Ising model from statistical physics is a powerful tool for modeling complex systems with interacting components. We apply the Ising model to the training of deep neural networks to explore the use of the Ising model for estimating the amount of data needed to train a certain architecture and finding universality in deep learning. One of the challenges in training deep neural networks is determining the optimal number of samples needed to achieve good performance. By using the Ising model, the research will attempt to estimate the minimum number of samples needed to train a given architecture. This will help to reduce the amount of time and resources needed for training, as well as improve the efficiency of deep learning algorithms. In addition, the research will explore the concept of universality in deep learning, which refers to the idea that certain properties of deep neural networks are independent of the specific architecture or dataset. The Ising model provides a framework for investigating universality in deep learning by modeling the interactions between neurons in a deep neural network. Overall, this research aims to contribute to the development of more efficient and effective deep learning algorithms by leveraging the Ising model from statistical physics. 

– ChatGPT