Analysing long time series of Earth observation images

Suggested by: Dirk Tiede, Martin Sudmanns

Background

Recent developments in satellite technology allow image acquisition with a high temporal frequency, yielding several tens of images per year for many geographical locations. Current approaches need to be adapted, e.g. for handling long Earth observation (EO) image time series. Data cubes are one solution to tackle this, one specific development are "semantic EO data cubes" with a first prototype of an operational Sentinel-2 semantic Earth observation (EO) data cube for Austria developed at Z_GIS.

Short description

A semantic EO data cube was defined as a spatio-temporal data cube containing EO data, where for each observation at least one nominal (i.e., categorical) interpretation is available and can be queried in the same instance (see Augustin et al, 2019).
This approach allows human users to query and analyse EO data on a higher semantic level (i.e. based on at least basic land cover units and encoded ontologies) directly in the data cube. A unique Web-interface allows interactive human-like queries based on spatio-temporal and semantic relationships. The new architecture allows a plethora of different approaches / query-models to be developed and applied to different use cases. Master theses could cover different topics making use of the generic semantic data cube for Austria (covering all Sentinel-2 data acquired) and may focus on the development of spatio-temporal queries for different use cases or on further development of parts of the system.

Suggested reading

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. Eur. J. Remote Sens. 50, 452–463.

Augustin, H., Sudmanns, M., Tiede, D., Lang, S., Baraldi, A., 2019. Semantic Earth Observation Data Cubes. Data 4. https://doi.org/10.3390/data4030102

Start

Anytime

Prerequisites/qualification

Interest in the topic++

Databases+

Programming or scripting languages++

Remote sensing++

Generating valid exposure data for bicycle crash analysis

Suggested by: Martin Loidl

Short description: Although lots of research on cycling safety is done, the results are of limited value in terms of risk assessment. This is mainly due to the absence of sound exposure data. Consequently, risk factors are only determined relative to the total number of crashes (x % of all crashes are caused by ...).
Different to motorized individual and public transport, transport models for cycling at the city scale are not well established yet. Additionally, available movement data are sparse. Stationary counting stations provide totals at a very specific location. Tracking data from mobile applications allow for a larger coverage, but suffer from selection biases. So far, a combination of stationary counts and mobile sensors has not been put into practice.
Thus, we are interested in

• a conceptual approach for relating stationary counting data with movement data (GPS trajectories)
• combining two different data sources for movement (GPS trajectories from a commercial application and number of traverses per road segment from a commercial fitness tracker) with stationary counting data in order to gain sound exposure data for subsequent crash analysis.

We can provide all data for the city region of Salzburg.

References, suggested reading:

• VANPARIJS, J., INT PANIS, L., MEEUSEN, R. & DE GEUS, B. 2015. Exposure measurement in bicycle safety analysis: A review of the literature. Accident Analysis & Prevention, 84, 9-19.
• DOZZA, M. 2017. Crash risk: How cycling flow can help explain crash data. Accident Analysis & Prevention, 105, 21-29.
• LOIDL, M., WALLENTIN, G., WENDEL, R. & ZAGEL, B. 2016. Mapping Bicycle Crash Risk Patterns on the Local Scale. Safety, 2, 17.

Related to projects: Bicycle Observatory (https://bicycle-observatory.zgis.at), SINUS (https://projekte.ffg.at/projekt/3325243)

Start/finish: ASAP


Prerequisites/qualifications: Profound skills in data modelling, management and processing as well as advanced skills in spatial (network-) analysis.

Spatio-temporal Patterns in Epidemiological Spread of Dengue and Chikungunya

Correlating User-generated Data with Public Health Data

Suggested by: Bernd Resch

Short description: Predicting epidemiological spreads in space and time is challenging because official health data is oftentimes not publicly available and partly outdated and often limited in its spatial and temporal resolution. Thus, this thesis will develop methods for identifying spatio-temporal patterns in dengue and chikungunya cases. The developed methodology will identify hot spots in space and time based on user-generated data (e.g., Twitter) and correlate the results with official and ancillary data (e.g. socio-economic & environmental data). Depending on the progress of the thesis, additional goals comprise the implementation of a tool for dynamic visualisation and the development of a method for forecasting the further spread of a disease, and investigating relationships with socio-economic information. The results can potentially support health institutions in planning vaccination and preparing for a rapidly spreading epidemic. The thesis has extraordinary practical relevance as chikungunya epidemic is currently spreading over South, Middle and North America, and Dengue fever is prevalent in many tropical and sub-tropical regions across the globe and increasingly impacts urban agglomerations.

The master thesis will be carried out together with Harvard University’s School of Public Health (HSPH) and the Center for Geographic Analysis (CGA).

Literature:

Hagenlocher, M., Delmelle, E., Casas, I., and Kienberger, S. (2013) Assessing socioeconomic vulnerability to dengue fever in Cali, Colombia: statistical vs expert-based modeling. International journal of health geographics, 12(1), 36.

Jaenisch, T. and Patz, J. (2002) Assessment of Associations between Climate and Infectious Diseases; a Comparison of the Reports of the Intergovernmental Panel on Climate Change (IPCC), the National Research Council (NRC), and United States Global Change Research Program (USGCRP). Global Change & Human Health 2002, Volume 3(1), 2-7.

Resch, B. (2013) People as Sensors and Collective Sensing - Contextual Observations Complementing Geo-Sensor Network Measurements. In: Krisp, J. (2013) Advances in Location-Based Services, ISBN 978-3-642-34202-8, Springer, Berlin Heidelberg, pp. 391-406.

Start date: ASAP

Prerequisites/qualifications: experience with analysing VGI, data visualisation, and spatio-temporal pattern detection

Identifying Urban Neighbourhoods from Social Media Posts

Urban Neighbourhood Definition beyond Administrative Boundaries

Suggested by: Bernd Resch

Short description: Urban neighbourhoods are inherently spatial and oftentimes defined through adminstrative districts. These neighbourhoods also represent urban places that are increasingly important units of focus in urban planning processes. Neighbourhoods are multi-dimensional in their nature, defined by physical, personal, political, organisational, socially enacted, and symbolic properties.
This thesis project aims to develop a new methodology for defining urban neighbourhoods through uncovering dynamic phenomena of pattern formation. Therefore, the analysis of social media posts has the potential of defining neighbourhoods through social interaction and interests, going beyond previous administrative definitions. This will lead to a new understanding of urban structures, potentially supporting urban planning through serving the needs of homogeneous communities with similar interests.

Literature:

Anselin, L. and Williams, S. (2016) Digital Neighborhoods. Journal of Urbanism: International Research on Placemaking and Urban Sustainability, 9(4), pp. 305-328.
Resch, B., Usländer, F. and Havas, C. (2018) Combining Machine-learning Topic Models and Spatiotemporal Analysis of Social Media Data for Disaster Footprint and Damage Assessment. Cartography and Geographic Information Science, 45(4), pp. 362-376.
Bernardo, F., & Palma-Oliveira, J. M. (2016) Urban Neighbourhoods and Intergroup Relations: The Importance of Place Identity. Journal of Environmental Psychology, 45, pp. 239-251.

Start date: ASAP

Prerequisites/qualifications: experience with geospatial analysis, data visualisation, and urban research

Spatial Simulation of a Skiing Resort

Suggested by: Dr. Christian Neuwirth (Z_GIS – Spatial Simulation) and Maximilian Mündler (PowerGIS GmbH)

Short description:

Winter tourism is a key player for Austria’s strong economic performance. Nowadays ski areas face different challenges such as climate change mitigation and rising competition among ski resorts. To stand these problems operating companies need to increase their efficiency.

In many ways, the infrastructure of a ski resort shows similarities to transportation networks. Lifts and slopes represent the network that skiers use for their activities. For efficiency enhancement of the operating infrastructure, simulations that respect the skiers’ behavior and the resulting consequences might be an interesting feature.

The aim of this thesis is to estimate the potential of spatial simulations as a tool for simulating the interactions between skiers and skiing infrastructure. Therefore, the typical behavior of skiers as well as the specifications of the infrastructure needs to be respected. Furthermore, the effects of punctual improvements as well as temporary changes (lift/slope closure) need to be investigated.

References, suggested reading:

POULHÈS, A. and MIRIAL, P. (2017): Dynaski, an Agent-based Model to Simulate Skiers in a Ski area. Procedia Computer Science 109: 84-91. DOI: 10.1016/j.procs.2017.05.298

Start/finish by: ASAP

Prerequisites/qualifications: Special interest in winter sport. Spatial Simulation knowledge.

Please contact Christian Neuwirth in case of interest: christian.neuwirth@sbg.ac.at