Topics for master thesis

• Analysis of Historical Changes in Farmlands using Landsat and Sentinel Satellite Data

  This master's thesis topic focuses on analysing changes in Estonian agricultural lands over a long-term perspective using satellite data. The goals are:

  Using Landsat data (from the 1980s onwards) and Sentinel data (from 2015 onwards), to map the distribution of arable lands during different time periods.

  To assess where and by how much farmlands have decreased or expanded, and to identify the main patterns.

  To try and find relationships between land use changes and, for example, protected areas, population density, or other landscape parameters.

  Within the scope of this work, you will learn to work with large volumes of satellite data and use cloud-based platforms like Google Earth Engine (GEE) for their analysis. You will acquire knowledge of the specifics of different types of satellite data (Landsat, Sentinel) and learn to classify them using machine learning methods. As a result of your work, you will produce a dataset that shows the changes in arable land over a long period.

  The topic is suitable for master's students specialising in geoinformatics. The topic assumes Python/R scripting/programming skills and a sufficient remote sensing background (LTOM.02.041 Geospatial Analysis with Python and R, and LTTO.00.027 Data Science in Remote Sensing).

• Innovating landscape metrics on hexagonal grids

Landscape metrics are widely used discovering patterns, changes, and trends in urban and rural landscapes. Landscape metrics is a well-developed field but hasn’t seen much progress lately. The aim is to develop and test implementation on hexagonal grids, where all six neighbours are equal, and compare that with well-established landscape metrics calculations that use classic raster 4 and 8 neighbourhoods. The topic is suitable for master students specialized in geoinformatics but are also interested in physical geography and landscape eceology. The topic requires Python/R scripting skills (LTOM.02.041 Geospatial Analysis with Python and R). Co-supervisor Evelyn Uuemaa.

• Developing remote sensing-based indicators for soil organic carbon ML modeling

• Towards analysis ready data cube for machine learning under FAIR open science

• A data observatory to explore multi-scale geographical relationships

The aim of this thesis topic is to apply spatial statistics and machine learning to a multi-resolution data cube, into which you will load various environmental, human-centric, and social-economic indicators. Overall, this effort is contributing to developing a Data Observatory for the European Data Spaces (https://digital-strategy.ec.europa.eu/en/policies/data-spaces) and Destination Earth mission. The focus will be on exploring relationships between variables of same and different resolution across local, regional, national and international scales.

The hypothesis is that at larger scales some relationships will be different than at small scales, while others stay stable, but don’t always know which and why. To navigate the modifiable areal unit problem (MAUP), there is also need to systematically evaluate different gridding strategies for the data. Results of the thesis will help to inform dataset and resolution choices for various spatial analysis tasks.

The topic is suitable for master students specialized in geoinformatics. The topic requires Python and SQL scripting and data management skills (LTOM.02.041 Geospatial Analysis with Python and R, and LTOM.02.040 Spatial Databases). Co-supervisor Evelyn Uuemaa.

• Advancing web cartography

Base maps are widely used in spatial data visualisation. However, the existing base maps for Estonia based on the Estonian Topographic Database have limited styling options and are only available in WMS. Alternatives in web map and mobile friendly tile-based (OSM XYZ / TMS) or Vector Tiles (MVT) format are only available from providers of Open Street Map data or commercial (Google, ESRI, o.a). The aim of this thesis is to improve on current efforts (see images) in web cartography at the Chair of Geoinformatics and Cartography of base maps styling based purely on Estonian authoritative data (Landboard). The task is to develop, improve, and describe styling methods for base maps in Geoserver, Mapnik, and MapBox VectorTiles for comparison. Special challenges include adequate labeling, label positioning and symbology.

This thesis is mostly focused on map design but also requires some engagement with web mapping technologies and the underlying spatial database, therefore the topic is only suitable for master students specialized in geoinformatics. Understanding of web mapping technologies is required (LTOM.02.043 Spatial Data Infrastructures / LTOM.02.067 Spatial Data on the Web). Familiarity with Postgresql/Postgis (LTOM.02.040 Spatial Databases) and Estonian geography in general would be highly desirable. Co-supervisor Raivo Aunap.

• What really are Story Maps?

The thesis will explore on a methodological and epistemological level what are “story maps” and what story maps aim to achieve. One challenge is to uncover how story maps make knowledge about geographical events more accessible. The topic is two-fold: a) requires developing several case studies on historical events, based on Estonian geo-referenced photo archives; b) various GIS methods and technologies are employed to draft a customized story mapping framework that aligns with the developed theoretical background, and does not rely on available proprietary or other pre-made software packages.

The topic is suitable for master students specialized in geoinformatics or Estonian Human Geography. Familiarity with Estonian language, geography and history in general is required. Background in at least one of a) web mapping technologies (LTOM.02.043 Spatial Data Infrastructures / LTOM.02.067 Spatial Data on the Web) or b) Python scripting skills (LTOM.02.041 Geospatial Analysis with Python and R) would be highly desirable. Co-supervisor Taavi Pae.

• Dynamics of heat urban islands in Tallinn
  
  Co-supervisor Dr. Mait Sepp, Senior Researcher in Physical Geography

Tallinn is a dynamic Estonian capital city whose outskirts are growing fast despite the economy drifting. With new developments, the Heat Urban Island (UHI) grows yearly. It is vital to communicate to the city planners the real effects of new developments: the location of hot spots, their shape and magnitude, their connection to different types of land use and available green space.
Fluent knowledge of geospatial analysis (vector and raster) is required; skills with remote sensing data are an asset. Possibility to candidate to Tallinn City scholarship.

• Learning Estonian crops’ phenological phases with remote sensing vegetation indexes.

Co-supervisors Ain Kull and Jaak Jaagus

A thorough knowledge of crops’ phenological phases is essential for the early detection of anomalies in plant development through the vegetation season. Also, due to climate change, local average phenology may drift from one decade to another. Remotely sensed vegetation indexes, such as NDVI, and their time series allow smooth phenological profiles to be created and phenologies and anomalies to be detected for different crops.

Fluent knowledge of geospatial analysis (vector and raster) is required; skills with remote sensing data are an asset.

Research fields: landscape elements, GeoAI

• Multi-class detection of small woody landscape elements using deep learning

In recent years, deep learning has been used to automatically map small woody landscape elements, such as hedgerows and field islands, from aerial and satellite imagery. Most of these studies have relied on convolutional neural networks (CNNs) that were pretrained on everyday photographs (e.g., the ImageNet dataset). While this approach has shown promise, it still struggles to distinguish between different types of small woody elements. This thesis will build on previous work by exploring whether models trained on remote sensing images (rather than natural images) can improve results. The project will focus on developing and testing a neural network that can perform multi-class segmentation, meaning it can separate different categories of small woody elements in agricultural areas. Publicly available orthophotos and existing landscape element maps will be used to train and evaluate the model. This project requires proficiency in the Python programming language and an interest in GeoAI. Co-supervisor Muhammad Afif Fauzan.

• How are different small landscape elements represented in the embedding space?

• Trends in drought frequency and geography in Estonia
• Diurnal weather cycles in Estonia
• Dynamics and geography of winter severity in Estonia
• Heat waves in Estonia

• Dissolved Organic Carbon (DOC) export and dynamics from managed peatland forests: the effect on inland water quality

Terrestrial and aquatic systems are interlinked; the nutrients and organic matter are transported in drainage water from land to lakes via streams and rivers. The changes in water quality also have drastic effects on aquatic biogeochemistry and ecosystem functioning. A large part of forests in Nordic countries are located on peatlands and since the demand for bioenergy is expanding rapidly, then there is a need for analysis of the environmental effects of peatland forest management practices on DOC and nutrient export from the soil. The aim of this thesis is to collect DOC data from forested peatland catchments and to analyse the effect of different forest management practices on the amount of DOC that is released to stream water.

• Effect of peatlands restoration to organic matter decomposition

• Investigating Time-Lagged Environmental Drivers of Methane Fluxes in Wetlands Using Eddy Covariance Data

• Linking Methane Flux Dynamics to NDVI Changes During the Vegetation Period

• The Dynamics of Sediment Accumulation in In-Ditch Constructed Wetlands

• Spatial Heterogeneity and Drivers of Methane and Nitrous Oxide Fluxes Across the Landscape

• Carbon and Nitrogen Balance of a Drained Peatland Forest Assessed with Eddy Covariance and Chamber-Based Measurements

• Efficiency of Subsurface Flow Constructed Wetlands in Removing Phenols from Wood Industry Leachate

• Evaluating Discrepancies in Annual Carbon Balance Using Eddy Covariance and Chamber-Based Measurements

• Length of growth period of pine and spruce according to dendrometers -

The dendrometers measure tree-trunk circumference every half an hour. Trunk circumference fluctuates diurnally. In the summer, a tree trunk grows thicker. The objective is to analyse the diurnal and annual patterns of tree growth.

• Ecosystem respiration in a hemiboreal pine forest in relationship with weather patterns - The thesis will be based on chamber measurements to be collected next year.
• Winter CO₂ fluxes in a hemiboreal pine forest - The material is the weather and CO2-concentration measurements collected in the Soontaga eddy tower.
• Literature analysis of the share of winter in the annual CO₂ balance
• Net ecosystem exchange of a hemiboreal pine forest - The material is the weather and CO₂-concentration measurements collected in the Soontaga eddy tower.

• Soil moisture and/or temperature derived from remote sensing in a freely chosen peatland of the world
• Vegetation and plant biodiversity in connection with climate, water regime and soil of a freely chosen peatland of the world
• Vegetation and plant biodiversity of a freely chosen peatland of the world derived from remote sensing
• Ecosystem services of the world’s mires and drained peatlands (e.g. clean freshwater, livable climate, food, biodiversity)
• Greenhouse-gas emissions of the world’s organic soils - The thesis will be based on the set of carbon dioxide (CO₂), methane (CH₄) and nitrous-oxide (N₂O) emissions, soil temperature, soil water content, soil chemistry, and plant traits collected from 70 sites all over the world’s organic soils by the University of Tartu geographers during 2011–2018.
• Impact of soil water and oxygen content on N₂O emissions - The thesis will be based on N₂O-emission and soil measurements from organic-soil sites in the vicinity of Tartu.
• Moisture content of the world’s organic soils - The objective of the thesis will be to predict moisture content of the world’s organic soils based on satellite data (e.g. Landsat). The predictions will be validated with the ground truth from the Department of Geography’s foreign expeditions.
• Nitrogen, phosphorus and organic-carbon runoff from catchments of the Baltic Sea - The thesis will be based on national stream monitoring data. It is possible to collect run additional measurements on a stream of the student’s choice.

• Meta-analysis of the effectiveness of environmental technological methods

The work includes work with scientific literature and data analysis. Please get in touch with me for further details.

Supervisor Mikk Espenberg, [email protected], homepage

• Microbial carbon and/or nitrogen cycle in different environments (forests, wetlands, artificial wetlands, farmland, etc.) and restored areas The work may include field work, laboratory work and data analysis. Please get in touch with me for further details.

Supervisor Mikk Espenberg, [email protected], homepage

• The relationship of microbial processes and microbial communities to greenhouse gases in different environments (forests, wetlands, artificial wetlands, farmland, etc.) and restored areas

The work may include field work, laboratory work and data analysis. Please get in touch with me for further details.

Supervisor Mikk Espenberg, [email protected], homepage

• Bioelectrochemical systems for water purification

The work may include field work, laboratory work and data analysis. Please get in touch with me for further details.

Supervisor Mikk Espenberg, [email protected], homepage

• Microbiological and isotope analysis methods to study biogeochemical cycles in the laboratory or in different environments (forests, wetlands, etc.)

The work may include field work, laboratory work and data analysis. Please get in touch with me for further details.

Supervisor Mikk Espenberg, [email protected], homepage

• Development, implementation and improvement of sustainable methods in different environments through the study of the microbial community

The work may include laboratory work and data analysis. Please get in touch with me for further details.

Supervisor Mikk Espenberg, [email protected], homepage

• Development of microbiological methods

The work may include field work, laboratory work and data analysis. Please get in touch with me for further details.

Supervisor Mikk Espenberg, [email protected], homepage

• Spatio-temporal dynamics of Land Surface Temperature (LST) and night frosts in Estonia

Co-Supervisor Dr. Valentina Sagris

Good knowledge of geospatial analysis (vector and raster) and natural geography is required.

• Detecting the vegetation dynamics of managed residual bogs using remote sensing

Co-Supervisor Dr. Valentina Sagris

Knowledge of spatial analysis (vector and raster data) and natural geography is required; Skills with remote sensing data are an asset.

• Overview of Estonian stormwater solutions and assessment of their performance

• Overview of nature-based stormwater maintenance plans

• Plant species composition of Viimsi raingarden and changes over time

• Public perception of nature-based stormwater solutions in Estonia

• Raingarden substrate composition and hydraulic properties

• Identify gaps in the design, implementation and maintenance of stormwater solutions affecting plant community and overall performance

• Residential segregation Tallinn

The aim is to analyze changes in residential distribution in the Tallinn city region by key population characteristics (income, ethnicity) since 1989 in longitudinal view. It is important to learn to use different segregation indices and methods for analyzing longitudinal data (e.g., event history analysis). The work is done within the framework of the chair’s research and under co-supervision.

• New immigrants in Estonia

The aim is to analyze new immigration to Estonia as a result of migration, number, origins and distribution in Estonia. It is important to learn how to use segregation indices and methods for analyzing longitudinal data (e.g., event history analysis). The work is done within the framework of the chair’s research and under co-supervision.

• School segregation in Estonia / Tallinn

The aim is to analyze changes in school segregation in the Tallinn city region since 1989 in a longitudinal view. It is important to learn to use different segregation indices and methods for analyzing longitudinal data (e.g., event history analysis). The work is done within the framework of the chair’s research and under co-supervision.

• Relationship between segregation and mobility

The aim is to analyze the residence and mobility of different population groups (gender, age, occupation, nationality) in the Tallinn city region. Learning how to use survey data and mobile positioning data as well as GIS skills are important. The work is done within the framework of the chair’s research and under co-supervision.

• New technologies for urban analysis

The aim is to familiarize yourself with the use of new technologies (drones, face recognition techniques), etc. in urban studies, with a focus on research on mobility and segregation. Curiosity and the desire to extend urban exploration to previously unknown territories are important. The work is done within the framework of the chair’s research and under co-supervision.

• Creating a 3D city model

The goal is to create a 3D city model for Tallinn and to equip houses with existing data, including Google's open data on dwellings, workplaces, schools and other important locations. Curiosity and the desire to extend urban exploration to previously unknown territories are important. The work is done within the framework of the department's research and under co-supervision.

• Estonian diaspora in the world

The aim is to explain the change in the number of Estonians in the world and the Estonians living in Finland in depth. Entrepreneurship in finding data in foreign communities and working with survey data is important. The work is done within the framework of the chair’s research and under co-supervision.

• Social media representations of urban neighbourhoods (e.g. based on locations and hashtags on Instagram) – comparing the images of neighbourhoods.

• Connecting social network to mobility - comparison of the spatial pattern of mobility and the location of call partners based on passive mobile positioning dataset.

• Demographic, social and ethnic dynamics in urban neighbourhoods. The role of spatial and social mobility. Ethnic differences in social and spatial mobility
• The impact of new residential construction on ethnic and social dynamics in urban neighbourhoods (the case of Tallinn urban region)
• Foreign migration trends. Residential, social and labor market integration of new migrants in Estonia
• Differences in activity spaces based on place of residence (differences by ethnicities) (applying mobile phone data)

• Social and ethnic differences in human mobility.

The objective is to find out what are the differences in the use of space and mobility of different social groups in Tallinn, and what factors influence it. Analysis is based on GPS accuracy MobilityLog app tracking data and questionnaire data from Tallinn.

• Social and ethnic differences in exposure to city functions.

An analysis of how urban functions (shops, services, etc.) are accessible to different social and ethnic groups throughout their activity space. Analysis is based on GPS accuracy MobilityLog app tracking data and questionnaire data from Tallinn.

• Modality styles and factors affecting these.

The objective is to find out madality styles and which social, place of residence, and urban space factors influence the people's travel modes usage (walking, cycling, public transport, car). Analysis is based on GPS accuracy MobilityLog app tracking data and questionnaire data from Tallinn.

• Spatio-temporal patterns of visiting second homes and other regular places.

Co-supervisor Ago Tominga

The objective of this study is to find out the distribution of second homes, temporal patterns of visiting them and indicators that influence it. Visiting patterns to second homes are an important source of information in regional planning. Second home is also an important factor in coping with disaster situations because second home offers an extra possibility to go when people are forced to evacuate from their first homes. The thesis develops a methodological framework to find second home visits based on mobile positioning data. Analysis is based on GPS accuracy MobilityLog app tracking data and questionnaire data.

• Regularity of mobility patterns of people.

Co-supervisor Ago Tominga

The objective of this study is to find out factors that influence the regularity of people’s mobility patterns. Indicators that describe people’s activity places and movement between them can be used to calculate regularity indicators of spatial mobility. Regularities with which activity space varies on different temporal scales (week, month, season) is an emerging theme in mobility research. The influence of social and personality indicators is estimated to the regularity measures. Analysis is based on GPS accuracy MobilityLog app tracking data and questionnaire data.

• Ethnic differences in mobility, domains of inter-ethnic contacts, neighbourhood change, ethnic identity and integration - analysis based on census, register and survey data.

• Gentrification

Gentrification is a process when younger and more affluent groups of society move into an area that was previously inhabited by people with lower incomes. As a result, the local real estate market revives, apartments are renovated, and often new community and leisure hotspots are created. The theory of gentrification points out how these developments can cause the revitalization of an area as well as the displacement of the long-term population, for example due to rapidly rising rents or the price of services. The aim of the thesis is to study residential and/or commercial aspects of gentrification process in post-industrial neighbourhoods and displacement pressure caused by revitalisation and gentrification.

• Effects of the digital innovation on creative and cultural industries

Creative activity has classically been seen as linked to a certain place: creative innovation has been seen to emerge in certain places (creative incubators, start-up clusters, etc.). You can also find creative campuses, incubators, co-working spaces in Estonia and elsewhere that benefit largely from geographical proximity. At the same time, the rapid development of technology has also changed the functioning of the creative and cultural industries. Many creative companies manage work and creative processes digitally and sell their goods and services online. The classically place-based creative economy is therefore at a very interesting crossroads: on the one hand, physical meeting places continue to be important, while the digital world is slowly taking over other activities. In Estonia, the effects of digitalization on creative entrepreneurship have not been studied at the master's thesis level. The goal is, for example, to investigate what are the effects of digitalization on creative business, what are the challenges and opportunities of digitalization, where are modern creative companies located? Suitable methods: interview, survey, spatial data analysis.

• Spatiotemporal patterns of train and bus use in Estonia based on ticketing data: input for mobility model validation

While public transit scheduling data has been open for a long time, train and regional bus ticketing and validation data provide new insights into the spatiotemporal patterns of public transit use across Estonia. The thesis could contribute to the LAAM mobility model project to support model validation and link public transit use to the sociodemographic composition of the Estonian population.

• The use of electric vehicles and accessibility to charging infrastructure across diverse neighbourhoods (together with Daiga Paršova)

In the V2G-QUESTS project, we investigate the opportunities and barriers for picking up electric mobility in large housing estate neighbourhoods, which have a diverse sociodemographic composition. In the thesis, it is possible to work with data from interviews, focus groups, a discrete choice modelling survey, and the traffic register to study behavioural choices, opportunities, and constraints further. When working with traffic register data, the student also contributes to the Estonian mobility model project.

• Cycling uptake in Baltic Sea cities (together with Veronika Mooses and Elise Jalonen)

Behavioural change from car use to sustainable urban mobility choices is the focus of the Cycle4Climate project. We have conducted a travel survey in four cities around the Baltic Sea to find out the potential of cycling interventions among diverse population groups. In the thesis, it is possible to study behavioural choices and cycling uptake across people with different travel profiles and preferences.

• Spatial planning systems of other countries

How does spatial planning work in other countries (possibly a comparison of countries or in-depth analysis of one specific country; also comparison with Estonia). What plans are being drawn up and why? What are the requirements for a planner? What could be suitable for Estonia from another country's system? "Planning families" of European countries.

• Spatial planning of objects with significant spatial impact (NIMBY objects)

How to build an effective planning process for an object that no one wants to see in their backyard? What are the successful and unsuccessful examples of large-scale planning in Finland and elsewhere? What are Estonia's opportunities and practice in compensating for the accompanying disturbances? What does planning a large building require from a planner, what from the community? The examples of Rail Baltic high-speed railway connection, wind farms and factories can be used.

• Spatial planning of green infrastructure

Changes in green infrastructure planning approaches, and specific plans for certain towns or rural municipalities.

• Local planning in municipalities

Master/comprehensive planning process in a town or rural municipality, the pros and cons of land-use planning, in-depth analysis of one certain comprehensive plan, comparison of comprehensive plans from different coutries.

• The role of the spatial planner

The changing role of the planner. Comparison of planner’s role in different countries or planning situations. What are the skills and knowledge required, and what are the perceptions over time? How is the training of the planner organized? What skills are lacking by planning practitioners?

• Participatory (green city) planning incl. impact assessment

Accessibility to electric vehicles among various neighbourhoods and socioeconomic population groups.

• How do varying household sociodemographic characteristics influence the likelihood of EV adoption in large-scale housing estates?
• What are the differences in EV-related infrastructure accessibility between newly built and older apartment districts?
• What are the most important facilitators (financial gain, energy independence) and concerns (battery degradation, control over charging times) for (potential) V2G adoption?