Suggested by: Martin Loidl

Short description: Parcel delivery systems increasingly rely on decentralised pickup stations whose performance and sustainability depend strongly on their spatial setting. Identifying suitable locations requires a detailed understanding of the spatial context and accessibility of potential locations.
This master thesis focuses on developing a spatial, data-driven approach to analyse and evaluate locations for parcel pickup stations. It includes compiling a comprehensive geodata inventory and applying accessibility modelling to assess site quality for walking, cycling, public transport, and motorised modes. Visual and analytic methods will be used to identify relationships between spatial structure, usage patterns, and CO2-reduction potential. The thesis may also involve developing or adapting automated GIS workflows for location scoring, or analysing the effect of different location types (e.g., residential areas, retail clusters, transportation hubs) on expected user behaviour.

By Matti Blume - Own work, CC BY-SA 4.0
https://commons.wikimedia.org/w/index.php?curid=69491925

Research for this master thesis should address one or more of the following questions (but is not limited to):

• Which spatial, infrastructural, and demographic characteristics define high-quality locations for parcel pickup stations?
• How can multimodal accessibility be modelled to support objective and comparable assessment of potential locations?
• How do different location types influence usage patterns and the potential reduction of motorised delivery trips?

References, suggested reading:

• PRANDTSTETTER, M., SERAGIOTTO, C., BRAITH, J., EITLER, S., ENNSER, B., HAUGER, G., HOHENECKER, N., SCHODL, R. & STEINBAUER, M. 2021. On the Impact of Open Parcel Lockers on Traffic. Sustainability, 13, 755. doi:10.3390/su13020755 
• NIEMEIJER, R. & BUIJS, P. 2023. A greener last mile: Analyzing the carbon emission impact of pickup points in last-mile parcel delivery. Renewable and Sustainable Energy Reviews, 186, 113630. doi:10.1016/j.rser.2023.113630 
• VAN DER MEER, L., WERNER, C. & LOIDL, M. 2024. Assessment of bicycle accessibility to mobility hubs under different criteria for cycling network quality. AGILE GIScience Ser., 5, 48. doi: 10.5194/agile-giss-5-48-2024
• OZYAVAS, P., BUIJS, P., URSAVAS, E. & TEUNTER, R. 2025. Designing a sustainable delivery network with parcel locker systems as collection and transfer points. Omega, 131, 103199. doi:10.1016/j.omega.2024.103199

Start/finish: anytime

Prerequisites/qualifications: Interest in spatial analysis, mobility behaviour, and data-driven urban logistics. Experience with GIS and data management is required; scripting skills (Python, R) are beneficial for automated workflows.

Spatial Group Model Building for Investigating Mobility Trends

Suggested by: Martin Loidl

Short description: Understanding mobility-related impacts requires modelling complex interactions between behaviour, infrastructure, and spatial context. Spatial Group Model Building (SGMB) offers a participatory method to conceptualize such systems by integrating stakeholder knowledge with spatial reasoning.
This master thesis investigates the suitability of SGMB for developing spatially explicit conceptual models, using the uptake of e-bikes among kids and young adolescents as a concrete use case. The work includes designing and applying an SGMB workflow, developing spatial causal diagrams, assessing methodological strengths and limitations, and exploring how SGMB outputs can inform subsequent quantitative analyses of mobility, safety, accessibility, or environmental impacts.

 Research for this master thesis should address one or more of the following questions (but is not limited to):

• How can SGMB support the development of spatially explicit conceptual models related to increasing e-bike use among kids and young adolescents?
• Which spatial, social, and infrastructural factors most strongly influence the likelihood that young adolescents adopt e-bikes, and how can these relationships be captured through Spatial Group Model Building?
• How do perceived risks (e.g., safety concerns), access to suitable infrastructure, and everyday mobility needs interact to shape e-bike uptake among kids and young adolescents in different spatial contexts?
• Which feedback mechanisms, such as changes in independent mobility, social norms, or exposure to traffic environments, drive or hinder the spread of e-bike use among adolescents, and how can SGMB help identify them across regions or settlement types? 

Suggested reading:

• SCOTT, R. J., CAVANA, R. Y. & CAMERON, D. 2016. Recent evidence on the effectiveness of group model building. European Journal of Operational Research, 249, 908-918. doi:10.1016/j.ejor.2015.06.078
• WEIR, H., BRENDAN, M., IRAKLIS, A., CLAIRE, C., CONOR, M., JOHN, B., ALBERTO, L., GARY, M., FRANK, K., RUTH, H. & AND GARCIA, L. 2024. Group model building for developing systems-oriented solutions to reduce car dependency in Belfast, United Kingdom. Cities & Health, 8, 374-389. doi:10.1080/23748834.2024.2328952
• Principles of group model building and spatial group model building: Slideshare 

Related project: This thesis can be linked to the i-MOBYL project. More information is available on the research group’s website.

Start/finish: anytime

Prerequisites/qualifications: Interest in mobility and transport research, participatory modelling, and spatial systems analysis. Experience with GIS is required; qualitative or conceptual modelling skills are an advantage.

 

Validating OpenStreetMap Data for Public Transport Stops Using Street-Level Imagery

Suggested by: Martin Loidl

Short description: Public transport (PT) operators and public authorities typically provide detailed location and timetable information on PT stops. However, data on the physical design, equipment, and immediate surroundings of these stops is often incomplete or entirely missing. This lack of information limits analyses on topics of high practical relevance, such as accessibility for users with disabilities, comfort and safety at stops, or environmental exposure (e.g., shade or sun).
OpenStreetMap (OSM) is widely recognized as a valuable source of transport-related geodata, offering a rich set of tags that describe physical features of PT stops. Yet, the completeness and accuracy of these attributes remain uncertain, particularly in rural areas.
This master thesis focuses on validating OSM data for public transport stops in rural contexts. It involves evaluating the accuracy and completeness of selected stop attributes (e.g., shelter, seating, signage, lighting) by systematically comparing OSM entries with street-level imagery from platforms such as Mapillary or Google Street View. The work includes developing a sampling strategy for rural stops, collecting and analyzing visual evidence, and documenting deviations between mapped information and observed reality. The goal is to assess OSM’s data quality for rural PT infrastructure and identify systematic patterns in missing or incorrect attributes.

The following research questions can guide the analysis:

• Which attributes of public transport stops are most frequently missing or incorrect in OSM?
• How reliable is street-level imagery for validating rural transport infrastructure?
• What sampling approach ensures representative coverage of rural stops?
• How can findings inform strategies for improving OSM data quality?

Related project: This thesis contributes to the SAFARI project, which focuses on identifying mobility barriers for vulnerable population groups. More information is available on the research group’s website.

Start/finish: anytime

Prerequisites/qualifications: Interest in mobility and transport planning, as well as in spatial data analysis. Skills in data management and geospatial analysis are required; scripting and coding skills are advantageous.

 

Modeling Place Vulnerability to Explosive Disease Outbreaks

Suggested by: Christian Neuwirth (Z_GIS – Spatial Simulation)
 
Short description:

In addition to the basic reproduction number, R0, the overdispersion parameter, k, plays a crucial role in characterizing the spread of infectious diseases. Estimates for COVID-19 indicate that the dispersion parameter k is approximately 0.1, suggesting that 80% of transmissions have been caused by only 10% of infectious individuals [1]. 

Observed overdispersion can arise from various factors. For instance, the same pathogen may exhibit different behaviors across individuals, e.g. the infectious period is better represented as a distribution rather than a fixed constant [2]. 

Additionally, observed overdispersion in disease transmission may stem from overdispersion in social contact networks. For example, a French social contact survey caried out by [3] demonstrated that a small number of individuals account for a disproportionately large share of overall social contacts, while many individuals have few or no social interactions.
Modeling experiments indicate that outbreaks within such social networks tend to be particularly explosive (Fig. 1).

 
Figure 1. The blue curves represent simulated outbreaks in empirical social networks exhibiting overdispersion, while the red curves depict outbreaks in networks where every individual has an equal number of social contacts. The basic reproduction numbers are as follows: R0=1.8 (A), R0=2.5 (B), R0=3.1 (C), and R0=3.7 (D).
 

Hypothesis: It is hypothesized that overdispersion in social contact networks is influenced by the physical structures of space, such as transportation infrastructure and other elements of the built environment. For instance, recent investigations showed that hierarchical cities are more vulnerable to the rapid spread of infectious diseases than decentralized cities [4]. In other words, overdispersion in physical structures translates into overdispersion in social contact networks, which in turn leads to overdispersion in disease transmission and explosive outbreaks.

The aim of this thesis is to quantify the vulnerability of locations to epidemic outbreaks by analyzing their structural properties.

Method: (1) Quantify the overdispersion parameter k in physical infrastructures using data from OpenStreetMap or open air travel network data (with the appropriate scale to be determined), (2) Run network simulations in a SIR-model (model is available) using the empirical parameter k as an input, (3) Compare epidemic doubling time in the simulation with empirical COVID-19 excess mortality doubling time at selected sites using a ranking scale approach.

Start: ASAP

Prerequisites/qualifications: 
Interest in spatial simulation and scripting (NetLogo, Python, R or GAMA)

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

References:

1. K. Sneppen, B. F. Nielsen, R. J. Taylor, and L. Simonsen, “Overdispersion in COVID-19 increases the effectiveness of limiting nonrepetitive contacts for transmission control,” Proceedings of the National Academy of Sciences, vol. 118, no. 14, p. e2016623118, 2021.
2. A. L. Lloyd, “Destabilization of epidemic models with the inclusion of realistic distributions of infectious periods,” Proceedings of the Royal Society of London. Series B: Biological Sciences, vol. 268, no. 1470, pp. 985–993, 2001.
3. G. Béraud et al., “The French connection: the first large population-based contact survey in France relevant for the spread of infectious diseases,” PloS one, vol. 10, no. 7, p. e0133203, 2015.
4. J. Aguilar et al., “Impact of urban structure on infectious disease spreading,” Scientific reports, vol. 12, no. 1, p. 3816, 2022.
5. O. Wegehaupt, A. Endo, and A. Vassall, “Superspreading, overdispersion and their implications in the SARS-CoV-2 (COVID-19) pandemic: a systematic review and meta-analysis of the literature,” BMC Public Health, vol. 23, no. 1, p. 1003, 2023.

Smallholder Farming and Global Crop Masks

Suggested by: Sophia Klaußner, Lorenz Wendt (CDL GEOHUM)

Short description: 
Studies have shown that landcover models are not sufficient in correctly identifying crop land in Sub-Saharan Africa. This has grave implications as early warning systems and the distribution of support in case of emergencies therefore gets significantly delayed.

In this study the student will investigate an area with small-holder farming to evaluate the accuracy of global models for small-holders and develop usability guidance from this. They will create a validation dataset and evaluate several global models for performance with small-scale agriculture and explore the implications related to this. Further they will create a help for deciding what land cover model to use in different contexts.

Suggested Reading:

• Dlamini, L., Crespo, O., Van Dam, J., & Kooistra, L. (2023). A global systematic review of improving crop model estimations by assimilating remote sensing data: Implications for small-scale agricultural systems. Remote Sensing, 15(16), 4066. https://doi.org/10.3390/rs15164066
• Kerner, H., Nakalembe, C., Yang, A., Zvonkov, I., McWeeny, R., Tseng, G., & Becker-Reshef, I. (2024). How accurate are existing land cover maps for agriculture in Sub-Saharan Africa? Scientific Data, 11(1), 486. https://doi.org/10.1038/s41597-024-03306-z
• Ketema, H., Wei, W., Legesse, A., Wolde, Z., Temesgen, H., Yimer, F., & Mamo, A. (2020). Quantifying smallholder farmers’ managed land use/land cover dynamics and its drivers in contrasting agro-ecological zones of the East African Rift. Global Ecology and Conservation, 21, e00898. https://doi.org/10.1016/j.gecco.2019.e00898

Interpretable Multimodal Machine Learning for Predicting and Explaining Livelihood Vulnerability to Drought in East Africa

Suggested By : Leizel De la Cruz, Lorenz Wendt

Objective:
To develop an interpretable machine learning framework for assessing livelihood vulnerability to drought in the arid regions of East Africa by integrating multimodal data (including Earth Observation, socio-economic and ancillary data) to enhance vulnerability prediction and identify the most influential underlying factors using SHapley Additive exPlanations (SHAP), thereby providing actionable insights for enhanced early warning systems.

Short Description:
Assessing livelihood vulnerability to drought is critical for proactive resilience-building in arid and semi-arid areas in East Africa. The traditional index-based methods usually depend on linear assumptions and expert-weighted indicators that might have overlooked the complex and non-linear relationships among factors contributing to vulnerability. This study proposes a data-driven approach that uses machine learning (ML) for prediction and the SHapley Additive exPlanations (SHAP) framework to explain and interpret the model output and identify the most influential predictors.

This research will first integrate different datasets including satellite drought indices; household survey data and other information related to livelihoods to develop a comprehensive dataset of predictors. Different ML models (like regression, XGBoost, random forest) will be trained on the historical vulnerability of livelihoods in district-level and identify the best model performance. The novelty of this research is in the application of SHAP analysis for an attempt to move beyond the "black box" nature of ML, as this approach quantifies the contribution of each identified factors (e.g. rainfall anomaly, SPEI, NDVI, market price fluctuation, distance to market) to the model’s ability to predict vulnerability. By making complex model outputs transparent and actionable, this research may provide a strong decision-support tool for enhancing drought resilience and decrease impacts to livelihoods in highly climate-sensitive regions.

Start: Anytime

Relevant Studies:

1. Crausbay, Shelley D., Kimberly R. Hall, Molly S. Cross, Meghan Halabisky, Imtiaz Rangwala, Jesse Anderson, and Ann Schwend. 2024. “A Flexible Data-Driven Approach to Co-Producing Drought Vulnerability Assessments.” Ecosphere 15(10): e70040. https://doi.org/10.1002/ecs2.70040
2. Enenkel, M., Steiner, C., Mistelbauer, T., Dorigo, W., Wagner, W., See, L., Atzberger, C., Schneider, S., & Rogenhofer, E. (2016). A Combined Satellite-Derived Drought Indicator to Support Humanitarian Aid Organizations. Remote Sensing, 8(4), 340. https://doi.org/10.3390/rs8040340
3. IPCC, 2022: Climate Change 2022: Impacts, Adaptation, and Vulnerability. Contribution of Working Group II to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change [H.-O. Pörtner, D.C. Roberts, M. Tignor, E.S. Poloczanska, K. Mintenbeck, A. Alegría, M. Craig, S. Langsdorf, S. Löschke, V. Möller, A. Okem, B. Rama (eds.)]. Cambridge University Press. Cambridge University Press, Cambridge, UK and New York, NY, USA, 3056 pp., doi:10.1017/9781009325844.
4. Lu, R., Liu, S., Duan, H., Kang, W., & Zhi, Y. (2024). Combining the SHAP Method and Machine Learning Algorithm for Desert Type Extraction and Change Analysis on the Qinghai–Tibetan Plateau. Remote Sensing, 16(23), 4414. https://doi.org/10.3390/rs16234414

Disaggregating Aggregated [Socioeconomic] Data into Grid-Level Representations Using PyInterpolate and Complementary Approaches for Humanitarian Applications

Suggested By: Khizer Zakir, Lorenz Wendt
 

Objective:
To investigate how geostatistical interpolation methods, as implemented in PyInterpolate, can be combined with complementary disaggregation approaches (dasymetric mapping, population weighting, machine learning-based downscaling) to transform aggregated socioeconomic data, such as demographics, education, health, or disease prevalence into fine-grained, grid-level representations. These representations may take the form of regular pixels or hexagonal H3 cells, depending on the resolution and structure most suitable for machine learning models. The thesis will quantitatively evaluate how such disaggregation improves the applicability and performance of ML models in humanitarian scenarios. In addition, this activity could include the uncertainties and explainability of such approaches.
 

Short Description:
Many critical datasets relevant to humanitarian decision-making, such as population demographics, education indicators, healthcare access, or disease spread are typically available only at coarse administrative levels (country, province, district). However, state-of-the-art machine learning models for spatial analysis generally operate on high-resolution gridded data, especially when integrating with environmental or remote sensing datasets. This mismatch in spatial resolution poses a barrier to building comprehensive, data-driven humanitarian models.

This thesis proposes to bridge this gap by studying the use of PyInterpolate and related interpolation/disaggregation systems to generate grid-level approximations of aggregated socioeconomic data. Both pixel grids and H3 hexagonal grids will be evaluated for their suitability in integrating heterogeneous datasets. The study will further assess the uncertainty of disaggregation outputs and their downstream impact on ML-based predictions.

Such an approach can be particularly important in humanitarian applications, where access to high-resolution socioeconomic data is scarce or delayed. Potential applications include:

• Disease spread modeling, where fine-scale integration of demographic and health data can improve outbreak prediction.
• Migration and human mobility studies, where disaggregated socioeconomic data can be compared and integrated with environmental drivers (floods, droughts, land degradation) at grid level to better understand displacement dynamics and population movements during crises.
• Disaster preparedness and response, where combining socioeconomic vulnerability layers with hazard data enables better risk assessments.
• Resource allocation and crisis monitoring, where timely, high-resolution information supports more equitable and effective humanitarian interventions.

 

Suggested Methodology:

• Geostatistical interpolation with PyInterpolate (kriging-based techniques).
• Covariate-driven disaggregation, using auxiliary layers such as land use, night-time lights, road networks, or population density as predictors of within-unit variation.
• Uncertainty quantification, with a particular focus on Bayesian approaches (Bayesian kriging, Bayesian hierarchical models, or probabilistic ML) to explicitly model uncertainty in disaggregation outputs and evaluate their downstream impact on ML-based predictions.

Start: Anytime

Relevant Studies:

1. Moliński, S., (2022). Pyinterpolate: Spatial interpolation in Python for point measurements and aggregated datasets. Journal of Open Source Software, 7(70), 2869, https://doi.org/10.21105/joss.02869 
2. Stevens, Forrest R., Andrea E. Gaughan, Catherine Linard, and Andrew J. Tatem., (2015). “Disaggregating Census Data for Population Mapping Using Random Forests with Remotely-Sensed and Ancillary Data.” PLoS ONE 10 (2): e0107042. https://doi.org/10.1371/journal.pone.0107042 
3. Wardrop, N. A., et al., (2018). “Spatially Disaggregated Population Estimates in the Absence of National Population and Housing Census Data.” Proceedings of the National Academy of Sciences 115 (14): 3529–37. https://doi.org/10.1073/pnas.1715305115