Semantic vs. Index-Based Bare Soil Mapping from Sentinel-2 Time Series: A Comparative Analysis Using Sen2Cube.at

 Suggested by: Dirk Tiede, Martin Sudmanns

Bare soil mosaic calculation based on semantic categorie in a semantic EO data cube

Short description

Mapping of bare soil exposure across agricultural landscapes is essential for monitoring soil erosion risk, estimating carbon stocks, and evaluating the effectiveness of soil conservation practices such as cover cropping. Reliable and temporally dense bare soil information furthermore provides the foundation for downstream tasks like soil organic carbon modelling and cropland management assessment at regional to national scales.

This thesis topic proposes to investigate whether semantic, knowledge-driven classification within the Sen2Cube.at Earth observation data cube framework can produce comparable bare soil composites from Sentinel-2 time series in contrast to conventional index- and threshold-based approaches. Rather than relying on normalised spectral indices such as NDVI or NBR2 - which lose absolute reflectance intensity, are influenced by clouds and may misclassify sparsely vegetated or mixed pixels as bare soil - the semantic approach encodes expert knowledge about surface conditions directly into the classification logic, potentially leveraging the full spectral profile of each observation. The resulting bare soil composites will be compared against established index-threshold products, such e.g. as the soil spectral suite developed at DLR, using both quantitative accuracy assessment against reference data and a qualitative analysis of how each method handles spectrally ambiguous surfaces. The comparison will specifically examine whether it produces more temporally consistent composites under varying conditions and different time periods.

Suggested reading

Sudmanns, M., Augustin, H., van der Meer, L., Baraldi, A., & Tiede, D. (2021). The Austrian semantic EO data cube infrastructure. Remote Sensing, 13(23), 4807. https://www.mdpi.com/2072-4292/13/23/4807

Heiden, U., d’Angelo, P., Schwind, P., Karlshöfer, P., Müller, R., Zepp, S., Wiesmeier, M., & Reinartz, P. (2022). Soil Reflectance Composites—Improved Thresholding and Performance Evaluation. Remote Sensing, 14(18), 4526. DOI: 10.1016/j.rse.2017.11.004

Start

As soon as possible

Prerequisites/qualification

• Remote Sensing & GIS
• Basic Scripting/Programming (e.g., Python)
• Interest in the topic 

Semantic analysis of gradual tree vitality loss using Sen2Cube.at

 Suggested by: Martin Sudmanns, Dirk Tiede

Abrupt changes in vegetation are easy to spot, gradual changes; however, such as those resulting from disease are difficult to identify.

Short description

Understanding the gradual loss of tree vitality is essential for sustainable forest management, particularly in the face of climate change and other environmental stressors. Long-term trends in forest health can be analysed using Earth observation data. Sen2Cube.at is a semantic EO data cube, that enables the analysis of large volumes of Sentinel-2 data semantically enriched over extended periods to monitor changes in forest vitality.

The goal of this master thesis is to develop a workflow for detecting and analysing temporal changes in forest vitality using the Sen2Cube.at . This study, to be conducted in collaboration with the “Bundesforschungszentrum für Wald” will focus on making use of spectral indices and semantic layers to assess vitality loss in forests, with a specific focus on gradual changes.

A specific forest area will be selected as a case study, with results validated using in-situ measurement. The outcomes of this thesis are expected to enhance the understanding of long-term forest health trends and support evidence-based forest management strategies.

Suggested reading

Sudmanns, M., Augustin, H., van der Meer, L., Baraldi, A., & Tiede, D. (2021). The Austrian semantic EO data cube infrastructure. Remote Sensing, 13(23), 4807. https://www.mdpi.com/2072-4292/13/23/4807

Start

As soon as possible

Prerequisites/qualification

• Remote Sensing & GIS
• Basic Scripting/Programming (e.g., Python)
• Interest in the topic 

Multiscale analysis of thermal data for urban sustainability

 Suggested by:  Dirk Tiede, Martin Sudmanns

Example data for a thermal acquisition during a hot summer day in a city

Short description

Understanding urban heat dynamics is critical for addressing challenges related to climate change, urban planning, and sustainability. Thermal data from various sources, including satellite imagery and aerial photographs, provide valuable insights into phenomena such as heat islands and green space changes. By analysing these data at multiple scales and times, it is possible to assess and monitor urban thermal behaviour more effectively.

The goal of this master thesis is to perform a multiscale analysis of thermal data as part of the funded project “Prometheus” – Progressive Methods of Thermal High-Resolution Earth Surveillance for Urban Sustainability. The study may focus on one or many of the following topics (tbd):

• Comparing day and night thermal data to analyze urban heat distribution patterns.
• Delineating urban heat islands based on thermal data pattern (spatial/temporal), urban structure and additional data integration
• Investigating the relationship between heat islands and changes in urban green spaces over time.
• Combining thermal data from different scales, such as high-resolution aerial imagery and satellite data, to enhance spatial and temporal understanding.

The outcomes are expected to contribute to sustainable urban planning strategies by providing actionable insights into heat management and green infrastructure optimization. The results will be documented with clear methodologies to support further applications in urban sustainability projects.

 Start

As soon as possible

Prerequisites/qualification

++ interest in the topic
+ programming skills
+ Earth observation

Change analyses: Soil sealing

 Suggested by: Martin Sudmanns, Dirk Tiede

Example of an analysis result that indicates vegetation loss or soil sealing.

Short description

The process of covering natural soil with impervious materials such as concrete or asphalt, in other words soil sealing, is a critical issue in urban planning and environmental management. Monitoring and analysing changes related to it over time is essential for sustainable land-use planning and for supporting authorities in decision-making processes.

This master thesis aims on analysing changes in soil sealing using Earth observation data and geospatial techniques. By employing advanced (geo)visualization methods, the study has the objective to provide intuitive and actionable insights into temporal patterns of soil sealing. The work will explore how these analyses can be effectively integrated into planning workflows and applied to support authorities in managing land-use changes. The results are expected to enhance urban planning strategies by offering tools for better understanding and mitigating the impacts of soil sealing on ecosystems and urban environments.

Suggested reading

Sudmanns, M., Augustin, H., van der Meer, L., Baraldi, A., & Tiede, D. (2021). The Austrian semantic EO data cube infrastructure. Remote Sensing, 13(23), 4807. https://www.mdpi.com/2072-4292/13/23/4807

Start

As soon as possible

Prerequisites/qualification

++ interest in the topic
++ programming skills
+ Earth observation

Live queries in the field: Bridging Earth Observation data and in-Situ Measurements

 Suggested by: Martin Sudmanns, Dirk Tiede

From big data to smartphone apps: Examples of Earth observation queries in the field.

Short description

Integrating Earth observation data with in-situ measurements for live analysis can be a powerful approach to enhance decision-making in agriculture, particularly within the framework of the Common Agricultural Policy (CAP). By enabling live queries in the field, accessing and analyzing EO data directly in real time become possible, supporting more accurate assessments of crop conditions, land use, and compliance with agricultural regulations.

This master thesis topic aims to explore the development of a seamless interface between EO data and in-situ measurements for agricultural applications. The objective is to create workflows that enable live queries in the field, utilizing EO datasets to provide actionable insights to stakeholders. Using a valuable tool such as Sen2Cube which is a semantically enriched data cube, the thesis will investigate how a mobile smartphone or tablet application can be used for live-interpretation of satellite imagery. The research outcomes are expected to improve the integration of EO data into agricultural monitoring systems, offering practical solutions for CAP-related assessments and supporting sustainable agricultural practices.

Suggested reading

Sudmanns, M., Augustin, H., van der Meer, L., Baraldi, A., & Tiede, D. (2021). The Austrian semantic EO data cube infrastructure. Remote Sensing, 13(23), 4807. https://www.mdpi.com/2072-4292/13/23/4807

Start

As soon as possible

Prerequisites/qualification

++ interest in the topic
++ programming skills
+ Earth observation

Evaluation and visualization of long earth observation time series

 Suggested by: Martin Sudmanns, Dirk Tiede

Examples of long Earth observation time series. Visualization capabilities are usually on a lower level than the analytical capabilities. 

Short description

Long-term Earth observation data provide valuable insights into environmental changes, particularly in sensitive regions like the Alps. Visualizing and analysing these data dynamically can help better understand temporal trends and support environmental monitoring. For example, Essential Climate Variables (ECVs), such as vegetation indices or snow cover, are key indicators for assessing the impacts of climate change in mountainous regions and require data for multiple decades.

 The objective of this thesis is to evaluate and visualize long EO time series using the Sen2Cube.at which is a semantic data cube that enables the analysis of large volumes of EO. The work can be summarized into the following key points:

• Developing dynamic visualizations of long EO time series to highlight temporal trends in selected ECVs for the Alpine region.
• Using Sen2Cube.at queries to efficiently extract and analyze semantic data related to ECVs, such as vegetation health or snow cover dynamics.
• Demonstrating the applicability of the approach through case studies of key climate variables in specific Alpine areas.

This master thesis workflow will provide innovative tools for environmental monitoring, offering dynamic and accessible visualizations of climate-related changes that can be used to support decision-making in the Alpine region.

Suggested reading

Sudmanns, M., Augustin, H., van der Meer, L., Baraldi, A., & Tiede, D. (2021). The Austrian semantic EO data cube infrastructure. Remote Sensing, 13(23), 4807. https://www.mdpi.com/2072-4292/13/23/4807

Start

As soon as possible

Prerequisites/qualification

++ interest in the topic
+ programming skills
+ Earth observation
+ Experience or willingness to get familiar with (3D) visualisation engines (Blender, unity, …)

Semi-Automated Delineation of Land Cover for Cartographic Use

 Suggested by: Johannes Scholz in cooperation with Dirk Tiede, Martin Sudmanns and the Österreichischer Alpenverein

Source: Österreichischer Alpenverein

Keywords: land cover delineation, cartographic representation, topographic maps

Objective: Create a reliable method to semi-automatically delineate land cover classes from base data for cartographic representation, reducing manual processing and enhancing map production quality.

Short Description: 

Geographic data is ubiquitous in the form of maps, though often presented poorly. To produce high-quality topographic maps, an exact delineation of land cover is required alongside many other processing steps. Distinguishing between forest and dwarf pines ("Latschen"), as well as between rocks and scree, still requires extensive manual work, especially when unsatisfied with poorly digitized OSM polygons. Reliably deriving these classes from available base data and preparing them for cartographic representation would add significant value to cartography.

Subsequent steps in map production, particularly the automated depiction of rocks and scree, rely on these delineations. A functional concept would also be welcomed by our partners, ensuring the practical application of the research results.

Start: Anytime

Big Earth observation analytics to support spatial planning

Suggested by: Martin Sudmanns, Hannah Augustin, Dirk Tiede

Left: Landslide in Bad Hofgastein with large impact on the landscape as seen by an aerial camera. Right: Sen2Cube.at analysis of three years of Sentinel-2 images before and after the event, the RGB composite is based on yearly vegetation layers, where the red colour indicates loss of vegetation after the year 2019 (= landslide extent)

Short description: The Sen2Cube.at System is an Earth observation (EO) data cube that allows cloud-based analyses on big EO data (focus on Copernicus Sentinel-2) and producing information that could be of interest for planning purposes and in the contexts of local governments. 

However, using EO data and derived products, particularly for custom, on-demand analysis is challenging and, in the context of local governments, is associated with several hurdles. This master thesis aims to investigate the use of Sen2Cube.at as a cloud-based system for creating custom EO analyses and using them in planning contexts of the local government. The tasks are: 

• Identifying the state-of-the-art and identifying the technical and organizational requirements 
• Developing a workflow for a cloud-based analysis of EO data using the Sen2Cube.at semantic EO data cube using one example (e.g. the landslide in Bad Hofgastein in 2020)  
• Identifying and prototypical development of interfaces into the local government’s workflows by considering their technical requirements previously defined 

The purpose of the master thesis is to create an end-to-end example of such a workflow using a concrete example and detailed documentation in German and English language. Expected is a good technical understanding and an understanding of the requirements and limitations of workflows in local governments. A collaboration with the local government in Salzburg is possible. 

Suggested reading:

Strasser, T., Sudmanns, M., Augustin, H., Van der Meer, L., Herzinger, K., Kerschbaumer, M., ... & Tiede, D. (2022). Identifying soil sealing hotspots on-demand for reporting and decision making in Austria using a Sentinel-2 based semantic EO data cube. In GI_Salzburg 2022. https://sims.sen2cube.at

Sudmanns, M., Augustin, H., van der Meer, L., Baraldi, A., & Tiede, D. (2021). The Austrian semantic EO data cube infrastructure. Remote Sensing, 13(23), 4807. https://www.mdpi.com/2072-4292/13/23/4807

Start/finish: ASAP

Prerequisites/qualifications: 

• Remote Sensing & GIS 
• Knowledge about spatial planning and local government workflows 
• Interest in the topic   
• German language skills 

Standardised automatic cross-sensor change detection in semantic Earth observation data cubes

 Suggested by: Dirk Tiede, Martin Sudmanns, Hannah Augustin

Short description: 

The Sen2Cube.at system facilitates the first instances of semantic Earth observation (EO) data and information cubes, e.g. using Sentinel-2 data captured over Austria and Syria, and AVHRR and Sentinel-3 data captured over the Alps. All of these semantic EO data cubes can be analysed via Web-based interfaces (e.g. via Web-browser). An EO data cube is a way of organising EO data that abstracts data storage so that users can access EO data based on spatio-temporal coordinates rather than their file names or directory structures, which makes it a lot easier to access the data. An EO data cube is considered a multi-dimensional structure with at least one non-spatial dimension (e.g., time), where coordinate tuples of the dimensions are used for data access. A semantically enriched EO data cube provides for each observation at least one nominal (i.e., categorical) interpretation, which can be queried in the same instance. 

The Sen2Cube.at system uses the satellite image automated mapper (SIAM) as semantic enrichment engine, which provides spectral categories from reflectance values. Using a per-pixel physical spectral model-based decision tree, SIAM automatically categorises EO imagery based on reflectance values from multiple optical sensors (e.g., Sentinel-2, Landsat-8, AVHRR, VHR). The software is capable of producing different granularities (i.e. different number of colour names) from coarse (i.e., 18 colour names) to fine (i.e., 96 colour names), as well as additional data-derived information layers (e.g., multi-spectral greenness index, brightness), fully automated without any specific parameterization.

The goal of this master thesis is to develop and implement a change matrix based on the spectral categories into the Sen2Cube.at system (i.e. transfer of the matrix into Sen2Cube.at models and/or Juypter notebooks) to allow fully automated change analysis for any image combination in the cube and can support various applications (drought, flood, snow cover change etc,). Selected applications can be demonstrated using Sentinel-2 data in Austria and Syria and AVHRR/Sentinel-3 data in the Alps. 

References, suggested reading:

• Sudmanns, M., Augustin, H., van der Meer, L., Baraldi, A., & Tiede, D. (2021). The Austrian Semantic EO Data Cube Infrastructure. Remote Sensing, 13(23), 4807. https://www.mdpi.com/2072-4292/13/23/4807
  Augustin, H., Sudmanns, M., Tiede, D., Lang, S., & Baraldi, A. (2019). Semantic Earth observation data cubes. Data, 4(3), 102. https://www.mdpi.com/2306-5729/4/3/102 

Related to projects: https://sen2cube.at 

Start/finish: anytime

Prerequisites/qualifications: 

    + Remote Sensing

    + Programming

Comparison of spectral categorisation of different Sentinel-2 bottom-of-atmosphere products

Suggested by: Martin Sudmanns, Thomas Strasser, Dirk Tiede

Short description: Satellite Earth observations (EO) are measurements: They need to be calibrated very well in order to provide high-quality information. One calibration step is the calculation of bottom-of-atmosphere reflectance values with the aim to remove or reduce atmospheric influences. Several different approaches and technical implementation exist in parallel (e.g. Sen2Cor, FORCE, ..)

The satellite image automated mapper (SIAM) provides spectral categories from reflectance values. Using a per-pixel physical spectral model-based decision tree, SIAM automatically categorises EO imagery based on reflectance values from multiple optical sensors (e.g., Sentinel-2, Landsat-8, AVHRR, VHR) The software is capable of producing different granularities (i.e. different number of colour names) from coarse (i.e., 18 colour names) to fine (i.e., 96 colour names), as well as additional data-derived information layers (e.g., multi-spectral greenness index, brightness). While SIAM-based categories are routinely derived from Sentinel-2 top-of-atmosphere reflectance values, they have not yet been used on bottom-of-atmosphere calibrated reflectance values.

The goal of this master thesis is to investigate and compare the spectral categories from SIAM on different bottom-of-atmosphere calibrated Sentinel-2 images.

References, suggested reading:

• Baraldi, A.; Durieux, L.; Simonetti, D.; Conchedda, G.; Holecz, F.; Blonda, P. Automatic Spectral-Rule-Based Preliminary Classification of Radiometrically Calibrated SPOT-4/-5/IRS, AVHRR/MSG, AATSR, IKONOS/QuickBird/OrbView/GeoEye, and DMC/SPOT-1/-2 Imagery—Part I: System Design and Implementation. IEEE Trans. Geosci. Remote Sens. 2010, 48, 1299–1325.

• Rumora, L.; Miler, M.; Medak, D. Impact of Various Atmospheric Corrections on Sentinel-2 Land Cover Classification Accuracy Using Machine Learning Classifiers. ISPRS Int. J. Geo-Inf. 2020, 9, 277. https://doi.org/10.3390/ijgi9040277

Related to projects: https://sen2cube.at 

Start/finish: anytime

Prerequisites/qualifications: 

    + Remote Sensing

    + Programming

Automated monitoring and alerting system based on EO image time series

Supervisors

Martin Sudmanns, Dirk Tiede

Background

Temporally high-frequency observations from Copernicus Earth observation (EO) satellites allow detecting and monitoring changes on the Earth’s surface. Changes may include short-term events (e.g. deforestation, flooding) or long-term trends and transitions (e.g. climate-change-induced vegetation changes). Earth observation data cubes are state-of-the-art infrastructure backbones to easier investigate the temporal dimension at scale. It is then possible to detect changes and produce information about types of changes retrospectively using existing time series data in the archives. The “live” monitoring based on continuously updated, new data (e.g. every few days for Sentinel-2 satellite images) in existing approaches are either limited to a specific application in the EO domain (e.g., for deforestation) or developed outside the EO domain and not yet applied and used in combination with EO data / EO data cubes (e.g. Grafana for monitoring IT systems).

Expected from the master thesis is an investigation of existing classifications of EO image time series changes and approaches to monitoring (natural) resources using EO data. Further, a generic method should be developed and (prototypically) implemented as a monitoring and alerting system based on frequently updated EO data cubes. This master thesis will be embedded into the overarching goal of building a semantic EO data cube infrastructure, which is developed at Z_GIS (https://sen2cube.at), and access to these data cubes will be provided.

Example dashboard based on Grafana for monitoring IT resources, including options to configure alerts for increasing, decreasing, or missing values.

Suggested reading

Hermosilla, T., Wulder, M. A., White, J. C., Coops, N. C., & Hobart, G. W. (2018). Disturbance-informed annual land cover classification maps of Canada's forested ecosystems for a 29-year landsat time series. Canadian Journal of Remote Sensing, 44(1), 67-87. https://www.tandfonline.com/doi/full/10.1080/07038992.2018.1437719

Kennedy, R., et al. Bringing an ecological view of change to Landsat‐based remote sensing. Frontiers in Ecology and the Environment 12.6 (2014): 339-346. https://esajournals.onlinelibrary.wiley.com/doi/full/10.1890/130066

Augustin, H., Sudmanns, M., Tiede, D., Lang, S., & Baraldi, A. (2019). Semantic Earth observation data cubes. Data, 4(3), 102. https://www.mdpi.com/2306-5729/4/3/102

Tiede, D., Baraldi, A., Sudmanns, M., Belgiu, M., & Lang, S. (2017). Architecture and prototypical implementation of a semantic querying system for big Earth observation image bases. European journal of remote sensing, 50(1), 452-463. https://www.tandfonline.com/doi/abs/10.1080/22797254.2017.1357432

Related projects

https://sen2cube.at

https://sims.sen2cube.at

Prerequisites/qualification

Remote sensing
Programming and databases