ALBERT

Description: In a warming Arctic, permafrost-related disturbances, such as retrogressive thaw slumps (RTS), are becoming more abundant and dynamic, with serious implications for permafrost stability and bio-geochemical cycles on local to regional scales. Despite recent advances in the field of earth observation, many of these have remained undetected as RTS are highly dynamic, small, and scattered across the remote permafrost region. Here, we assessed the potential strengths and limitations of using deep learning for the automatic segmentation of RTS using PlanetScope satellite imagery, ArcticDEM and auxiliary datasets. We analyzed the transferability and potential for pan-Arctic upscaling and regional cross-validation, with independent training and validation regions, in six different thaw slump-affected regions in Canada and Russia. We further tested state-of-the-art model architectures (UNet, UNet++, DeepLabv3) and encoder networks to find optimal model configurations for potential upscaling to continental scales. The best deep learning models achieved mixed results from good to very good agreement in four of the six regions (maxIoU: 0.39 to 0.58; Lena River, Horton Delta, Herschel Island, Kolguev Island), while they failed in two regions (Banks Island, Tuktoyaktuk). Of the tested architectures, UNet++ performed the best. The large variance in regional performance highlights the requirement for a sufficient quantity, quality and spatial variability in the training data used for segmenting RTS across diverse permafrost landscapes, in varying environmental conditions. With our highly automated and configurable workflow, we see great potential for the transfer to active RTS clusters (e.g., Peel Plateau) and upscaling to much larger regions.

Description: “Artificial Intelligence for Cold Regions” (AI-CORE) is a collaborative project of the German Aerospace Center (DLR), the Alfred Wegener Institute (AWI), the Technical University Dresden (TU Dresden), and is funded by the Helmholtz Foundation since early 2020. The project aims at developing artificial intelligence methods for addressing some of the most challenging research questions in remote sensing of the cryosphere. Rapidly changing ice sheets and thawing permafrost are big societal challenges, hence quantifying these changes and understanding the mechanisms are of major importance. Given the vast extent of polar regions and the availability of exponentially increasing satellite remote sensing data, intelligent data analysis is urgently required to exploit the full information in satellite time series. This is where AI-CORE comes into play: Four geoscientific use cases have been defined, including a) change pattern identification of outlet glaciers in Greenland; b) object identification in permafrost areas; c) edge detection of calving fronts of glaciers/ice shelves in Antarctica; d) firn line detection and monitoring: The glacier mass balance indicator. For these four use cases, AI-methods are being developed to allow for an accurate, efficient, and automated extraction of the desired parameters. Once these methods have been successfully developed, they will be implemented in processing infrastructures at AWI, TU Dresden, and DLR, and subsequently made available to other research institutes. The presentation will outline the specific goals and challenges of the four use cases as well as the current state of the developments and preliminary results.

Description: Retrogressive thaw slumps (RTS) are typical landforms indicating processes of rapid thawing and degrading permafrost. Their abundance is increasing in many regions and quantifying their dynamics is of high importance for assessing geomorphic, hydrologic, and biogeochemical impacts of climate change in the Arctic. Here we present a deep-learning (DL) based semantic segmentation framework to detect RTS, using high-resolution multi-spectral PlanetScope, topographic (ArcticDEM elevation and slope), and medium-resolution multi-temporal Landsat Trend data. We created a highly automated processing pipeline, which is designed to allow reproducible results and to be flexible for multiple input data types. The processing workflow is based on the pytorch deep-learning framework and includes a variety of different segmentation architectures (UNet, UNet++, DeepLabV3), backbones and includes common data transformation techniques such as augmentation or normalization. We tested (training, validation) our DL based model in six different regions of 100 to 300 km² size across Canada, and Siberia. We performed a regional cross-validation (5 regions training, 1 region validation) to test the spatial robustness and transferability of the algorithm. Furthermore, we tested different architectures, backbones and loss-functions to identify the best performing and most robust parameter sets. For training the models we created a database of manually digitized and validated RTS polygons. The resulting model performance varied strongly between different regions with maximum Intersection over Union (IoU) scores between 0.15 and 0.58. The strong regional variation emphasizes the need for sufficiently large training data, which is representative of the diversity of RTS types. However, the creation of good training data proved to be challenging due to the fuzzy definition and delineation of RTS. We are further continuing to improve the usability and the functionality to add further datasets and classes. We will show first results from the upscaling beyond small test areas towards large spatial clusters of extensive RTS presence e.g. Peel Plateau in NW Canada.

Description: Retrogressive thaw slumps (RTS) are typical landscape processes of thawing and degrading permafrost. To this point, their distribution and dynamics are almost completely undocumented across many regions in the permafrost domain, partially due to the lack of data and monitoring techniques in the past. We are tackling this shortcoming by creating a deep learning based semantic segmentation framework to detect RTS, using multi-spectral PlanetScope, derived topographic (ArcticDEM) and multi-temporal Landsat Trend data. We created a highly automated processing pipeline, which is designed to create reproducible results and to be flexible for multiple input features. The processing workflow is based on the pytorch deep-learning framework and includes a variety of different segmentation architectures (UNet, UNet++, DeepLabV3), backbones and includes common data transformation techniques such as augmentation or normalization. We tested (training, validation) our DL based model in six different regions of 100 to 300 km² size across Canada (Banks Island, Tuktoyaktuk, Horton, Herschel Is.), and Siberia (Kolguev, Lena). We performed a regional cross-validation (5 regions training, 1 region validation) to test the spatial robustness and transferability of the algorithm. Furthermore, we tested different architectures backbones and loss-function to identify the best performing and most robust parameter sets. For training the models we created a training database of manually digitized and validated RTS polygons. The resulting model performance varied strongly between different regions with maximum Intersection over Union (IoU) scores between 0.15 and 0.58. The strong regional variation emphasizes the need for sufficiently large training data, which is representative for the massive variety of RTS. However, the creation of good training data proved to be challenging due to the fuzzy definition and delineation of RTS, particularly on the lower part. We have recently expanded our analysis to several RTS-rich regions across the Arctic (Fig.X) for the year 2021 and annual analysis (2018-2021) for RTS hot-spots, e.g. Banks Island, Peel Plateau and others. First model inference runs are promising for detecting RTS, but are still strongly overestimating the number and area of RTS, due to an excessive number of false positives. Model performance however, varies strongly between regions. Due to the strong variability of landscapes with RTS, we expect an improvement in model performance with an increase in the number and spatial distribution of training datasets. The community driven formation of the IPA Action Group RTSIn, which aims to create standardized RTS digitization protocols and training datasets for deep/machine-learning purposes will be a great boost for our purpose. With our standardized processing pipeline (preprocessing, training, inference), which allows to add more features based on user interest and data availability,, we tested our workflow for surface water and pingos with a mixture of publically available (Jones et al) and digitized data (Grosse pingos, Nitze water). These tests produced very good results and showed that the designed workflow is transferrable beyond the segmentation of RTS only. In the near future, we are aiming to integrate the community based training data and further gradually improve our training database. Within the framework of the ML4Earth project, we will create a temporal and pan-arctic monitoring system for RTS based on our highly automated processing chain. This will enable us to better understand pan-arctic RTS dynamics, their influencing factors, and consequences. Combining these spatial-temporal datasets with volumetric change information and carbon stock information will enable us to better quantify the consequences of thaw slumping across the permafrost domain.

Description: Artificial Intelligence for Cold Regions (AI-CORE) is a collaborative project of the DLR, the AWI, the TU Dresden, and is funded by the Helmholtz Foundation since early 2020. The project aims at developing AI methods for addressing some of the most challenging research questions in cryosphere remote sensing, rapidly changing ice sheets and thawing permafrost. We apply data analytics approaches to discover the data variable from data set simulated with an ice sheet model, observe the migration, and time Series analysis to predict and contrast this to simulated grounding line position. For the data assimilation in simulations of the Greenland ice sheet, we engage a level set method, that allows to derive a continuous function in time and space from discrete information at satellite acquisition time steps. We use an alpha-shape method to derive a seamless product of the margin at each time step to be used in the level set method driving the simulations. We develop AI algorithms and tools that allow scaling of our analyses to very large regions. Here we focus on the detection of Retrogressive Thaw Slumps (RTS), highly dynamic erosion processes caused by rapid permafrost thaw. We apply deep-learning based object detection on dense time-series of high-resolution (3m) multi-spectral PlanetScope satellite images and auxiliary datasets such as digital elevation models. RTS detection is challenging, as they are difficult to define semantically and spatially and are highly dynamic and embedded in different landscape settings. The results will help to understand, quantify and predict RTS dynamics and their landscape-scale impacts in a rapidly warming Arctic. We upgrade the base IT-infrastructure at AWI by integrating new GPU computing hardware into the on-premise IT-infrastructure to speed up the computing, data storage capabilities, and parallel processing, supporting the analytical workflows specifically.

Description: In a warming Arctic, permafrost-related disturbances, such as retrogressive thaw slumps (RTS), are becoming more abundant and dynamic, with serious implications for permafrost stability and bio-geochemical cycles on local to regional scales. Despite recent advances in the field of earth observation, many of these have remained undetected as RTS are highly dynamic, small, and scattered across the remote permafrost region. Here, we assessed the potential strengths and limitations of using deep learning for the automatic segmentation of RTS using PlanetScope satellite imagery, ArcticDEM and auxiliary datasets. We analyzed the transferability and potential for pan- Arctic upscaling and regional cross-validation, with independent training and validation regions, in six different thaw slump-affected regions in Canada and Russia. We further tested state-ofthe- art model architectures (UNet, UNet++, DeepLabv3) and encoder networks to find optimal model configurations for potential upscaling to continental scales. The best deep learning models achieved mixed results from good to very good agreement in four of the six regions (maxIoU: 0.39 to 0.58; Lena River, Horton Delta, Herschel Island, Kolguev Island), while they failed in two regions (Banks Island, Tuktoyaktuk). Of the tested architectures, UNet++ performed the best. The large variance in regional performance highlights the requirement for a sufficient quantity, quality and spatial variability in the training data used for segmenting RTS across diverse permafrost landscapes, in varying environmental conditions. With our highly automated and configurable workflow, we see great potential for the transfer to active RTS clusters (e.g., Peel Plateau) and upscaling to much larger regions.