Open calls in doctoral studies

The main application period took place from 1 to 15 May 2022.

Doctoral students admitted to these projects will have employment contracts with the junior researcher. The expected workload is 1.0, the expected work period is four years.  The final workload will be set during the negotiations.

Admission conditions and evaluation criteria of the faculty

In the Faculty of Science and Technology, all candidates must submit a motivation letter and a CV in the Dreamapply together with the application. Candidates will be assessed on the basis of a motivation letter and an entrance interview. (except for the science education, where a draft for doctoral project must be submitted instead of a motivation letter). Candidates will apply for announced projects.

Biodiversity and Ecological Sustainability

Supervisors: Tuul Sepp, Lauri Saks

Freshwater habitats have been facing threats of increasing severity, including habitat loss, pollution, fragmentation of rivers by dams, and novel diseases. Resulting from these and other from human activities, freshwater biodiversity faces unprecedented threats. To be able to stop the degradation of freshwater habitats, it is crucial to gather strong scientific evidence to indicate the link between human activities and wild animal diversity and welfare, but also to support the application of mitigation strategies. As of now, there are considerable research gaps in this field of study. The current doctoral project is built on ongoing habitat restoration projects, focusing on physiological condition indices of animals living in degraded and/or restored habitats, and on the effects of habitat connectivity on genetic diversity. Physiological metrics and genetic tools can help to detect the causes of declines in biodiversity, but applying these methods requires integration of the fields of conservation physiology, restoration genetics, and aquatic ecology. This project aims to fulfill this goal, to provide high quality scientific support for ongoing and future conservation efforts in freshwater habitats. The project focuses on following anthropogenic factors affecting freshwater habitat: (1) damming, and different strategies for creating bypasses, (2) urbanization, and the effect of landscape maintenance regime on urban stream fauna, (3) chemical pollution levels in natural rivers and streams. By assessing physiological condition indices and genetic diversity of fish from natural, degraded, and restored habitats, we aim to provide better tools for designing effective restoration projects.

Supervisor: Aveliina Helm

In recent years, there has been a growing concern regarding the global loss of biodiversity whilst several prestigious reports have assessed the problem (IPBES 2019; Lambertini 2020). Habitat loss and degradation, landscape and climate change, the spread of invasive species, and excessive use of natural resources have been identified as the main causes of the decline. One of the conclusions of the reports has been that in order to reduce the loss of biodiversity the landscapes must be purposefully managed to find ways to maintain biodiversity and provide resources needed by humanity at the same time.

One of the factors affecting biodiversity in our landscapes is the development of wind generators, to produce electricity from a renewable source of energy. The downside of wind energy is that the generators have an impact on biodiversity and one of the impacted taxa, already undergoing a decline, is bats (Chiroptera) (Kunz et al. 2007). The aim of the doctoral thesis is to evaluate the possible ways of deploying wind generators to reduce their impact on bats and related taxa. Scientific studies in the thesis are aimed at gaining knowledge to consolidate the needs of humans with biodiversity. The work has significant scientific and applied outcomes, delivering knowledge on understudied species and enabling us to manage populations of taxa susceptible to anthropogenic pressures.

Supervisor: Carlos Perez Carmona

From tiny duckweeds to gigantic sequoias, the extent of trait variation among vascular plants on Earth is extraordinary. Despite all this functional diversity, species’ ecological strategies resulting from trait combinations are constrained by physiological limits set by evolutionary history and trade-offs in resource allocation. Aiming to understand what are the main dimensions of functional variation and how species are organized within them, ecologists have recently started mapping the functional spectra of different taxonomic groups. These characterizations in the case of plants have only considered aboveground traits so far, despite the importance of belowground traits for plant biology. This project aims to fully incorporate belowground traits into the global spectrum of plant form and function, and to map how different plant communities and species are organized within this space. We will use this new knowledge to try to predict the effects of projected climate change in the functional structure and associated ecosystem functioning of plant communities at a global scale.

Supervisor: Kadri Põldmaa

Numerous hotspots of biological diversity and endemics are incorporated in various types of forests in South America. However, large part of the biodiversity yet remains to be discovered and described. The planned study focuses on the poorly studied guild of fungicolous fungi frequently infecting fruitbodies of mushrooms and polypores, while producing a variety of secondary metabolites. The aim of the proposed project is to elucidate the diversity and host associations of fungicolous fungi in the neotropics, a part of which we hypothesize to be restricted to this region like reported for the potential hosts. For that purpose, such fungi will be collected in different types of forests in Colombia, Brazil and Argentina, and allocated for morphological characterisation and DNA barcoding. Cryptic taxa will be screened by assigning ITS rDNA sequences to UNITE Species Hypotheses, that will reveal their host and geographic ranges. Constructing multigene phylogenies of the family Hypocreaceae allows to map the host associations and to reveal relationships of species in the neotropics.

The project is expected to result in papers focusing on coevolutionary patterns and cryptic diversity among Sepedonium species inhabiting fruitbodies of boletes as well as host specialization and distribution of fungicolous Hypocreaceae on basidiomycetes with agaricoid, polyporoid and corticioid fruitbodies. The innovativeness of combining traditional and modern methods of taxonomy will offer a holistic picture of the diversity, phylogeny, host interactions and geographic distribution of studied taxa. This will be achieved by analyzing all globally available DNA barcodes from diverse environments via the UNITE SHs that will be linked to phylogenetic trees constructed using multigene data from morphologically characterized voucher specimens. DNA barcoding of fungicolous fungi and their hosts will allow a better understanding of (cryptic) diversity and interactions among the two partners and facilitate identification of neotropical fungi in future metabarcoding studies.

Supervisors: Marina Semchenko, Mari Moora

Many ecosystems exhibit alternative stable states, such as vegetation dominated by woody or herbaceous species, that cannot be explained by climatic factors. The probability of a location being dominated by woody or herbaceous vegetation may be driven by local perturbations such as herbivory or fire. Evidence is accumulating that soil microbial communities could play a major role in structuring plant communities. However, their role in driving transitions between alternative stable vegetation states remains poorly understood. This PhD project aims at disentangling the microbial drivers of shifts between grassland and forest vegetation and the role of microbial communities in stabilising grassland vegetation in the face of external perturbations, such as an increasing frequency of severe droughts driven by climate change. The project will focus on different types of mutualistic associations between plants and mycorrhizal fungi as key drivers of transition between grassland and forest. In addition, the project will investigate the consequences of soil biodiversity loss in agricultural landscapes, focussing on the role of mutualist limitation in grassland restoration and resilience to drought events. Collectively, this project will advance our understanding of the roles microbial communities play in maintaining vegetation at a stable state and will provide recommendations for effective habitat restoration via microbial manipulation.

Supervisors: Triin Reitalu, Aveliina Helm

The high species diversity of semi-natural grasslands is resulting from long-term traditional extensive management. The ceasing of management during the last century has led to drastic decreases in the area of semi-natural grasslands and in their biodiversity. Restoration of semi-natural grasslands is, therefore, of high priority, especially in the face of ongoing climate and biodiversity crises. Traditionally, restoration success is monitored by following the recovery of one or two organism groups. However, the knowledge about how the interactions between different organism groups recover after the restoration and what are the potential factors hindering the full recovery of ecosystems is essential for planning future restoration projects.

Current PhD project takes advantage of the existing monitoring scheme with permanent study sites that was set up in 2015 to follow the restoration success of dry calcareous grasslands in Estonia. The project will build on existing data about the environmental conditions, abundance and diversity of vascular plants, mosses, lichens, soil fungi, bumblebees, butterflies, birds, spiders and Myriapoda from before and 2-3 years after the restoration. During the project, the permanent study sites are revisited and chosen groups of organisms are monitored again. In addition, more specialised species-species interactions (for example, between a plant species and its pollinator) are identified from existing data and studied in more detail.

The overall aim of the PhD project is to clarify what hinders the recovery of interactions between different organism groups in restored semi-natural grasslands and which landscape and historical factors influence the speed of recovery.

Supervisor: Tiit Teder

The ongoing loss and fragmentation of seminatural grasslands exert strong pressure on the biodiversity of open habitats in temperate environments. Open seminatural habitats are increasingly limited to linear landscape elements, such as road verges, powerline corridors and ditch banks. However, regarding insect biodiversity, little is known about the potential of such linear features to mitigate the negative effects of grassland loss. Taking maximum advantage of linear features requires good knowledge of the relative roles of local, landscape and management effects in the build-up of local species assemblages. Moreover, to adequately address this landscape planning challenge, we also need to identify the minimal proportion of seminatural habitats required in the landscape to prevent regional loss of biodiversity. In the case of insects, however, establishing such thresholds requires much larger data sets than have been available to date.

In this doctoral project, we address these knowledge gaps by running detailed analyses of butterfly species richness and composition over gradients of habitat and landscape variables, with the main focus on linear landscape elements. For our objectives, we primarily rely on two large data sets of butterfly surveys, collected as a result of the Estonian Butterfly Distribution Mapping and accumulating from the National Butterfly Monitoring Scheme. We complement these data sets by i) field- and GIS-based data on habitat, landscape and management variables, and ii) literature-derived phylogenetic and ecological data (body size, phenology, larval diet breadth, mobility) on butterflies. These three layers of data allows us to link variations in habitat and landscape parameters with particular species traits in a phylogenetically explicit framework.

Supervisors: Kessy Abarenkov, Otto Miettinen

The main purpose of this project is to extract phylogenetic and evolutionary information from existing data including published genomes and extensive unpublished raw data. The PhD student will develop a methodology for incorporating lower quality genomic data to analyses and will update the phylogeny of the classes Agaricomycetes and Dacrymycetes. The evolutionary transitions between various wood-decay mechanisms will be studied in detail.

The resulting robust phylogenetic tree is necessary for biodiversity studies. Functional annotation of genomes will provide information for further use of wood-decaying fungi in biotechnology such as pulp or plastic waste management.

Supervisors: Jane Oja, Sten Anslan

Indoor microbes play an important role in our daily lives and baseline information about them is therefore essential for identification of tentative health risks and, if necessary, for improved indoor air quality. The recent development of high-throughput sequencing methods has allowed quick and more thorough characterization of different microbial communities from various ecosystems, as well as from buildings, and given a better understanding of the factors that shape these communities. The main aim of this PhD thesis is to study the temporal and spatial patterns of the diversity of indoor fungal and arthropod communities. In particular, we ask how fungal and arthropod communities change along latitudinal gradients, and what kind of changes happen at the local scale during a period of one and a half years. In addition, we will detect potentially harmful microbes, their distribution, and factors that shape this distribution. This PhD thesis relies on samples, which have already been collected in the frameworks of “FunHome” citizen science campaign and microbial communities will be described by using up-to date molecular tools.

Supervisor: Prof. Leho Tedersoo

The main purpose of this project is to elaborate the conception of biodiversity crediting that would alternate or complement to the existing carbon crediting system. Using Estonian agricultural and forest lands as models, the PhD student will perform cost and benefit analysis of biodiversity crediting and scale it up to the entire Europe. Soil microbial and animal diversity will be used as a proxy for ecosystem biodiversity due to its economic feasibility and scalability. The PhD student will collect samples from Estonia and rest of the Europe. Within the project, economic cost-benefit analysis, conservation biology, microbiology and bioinformatics are integrated. Results of the study have a potential to maintain global biodiversity at current critical era and lead creation of a novel multi-billion business. The PhD candidate will be employed by the university of Tartu (as junior researcher) and EcoBase OÜ.


Supervisors: Enn Lust, Jaak Nerut, Vitali Grozovski and Anthony Kucernak

Electrochemical devices such as fuel cells and electrolysis cells provide an alternative to fossil fuel-based energy storage, chemical feedstock, and means of transportation. Fuel and electrolysis cells use catalysts to increase efficiency and selectivity. Platinum is the most widely used catalyst due to its high activity and stability. However, the cost of platinum contributes significantly to the overall cost of fuel and electrolysis cells. A better understanding of the reaction mechanisms on platinum at high current densities could guide the design of alternative lower-cost catalysts. The rotating disk electrode (RDE) is essential in investigating these mechanisms but suffers from slow mass transport. Consequently, RDE has a limited potential range for characterisation, and the results can differ from measurements in actual cells.

An alternative, called the floating electrode (FE) technique, provides much better mass transport conditions and can be a more reliable method for catalyst characterisation. The developers at Prof. A. Kucernak's group have used the FE to model hydrogen oxidation and evolution reaction (HOR/HER) and oxygen reduction and evolution reaction (ORR/OER) in acidic conditions. On the other hand, HOR/HER and ORR/OER have not been studied in alkaline high mass transport conditions, which would benefit alkaline fuel and electrolysis cell research. This project aims to develop an alkali-compatible FE technique and investigate the electrochemical reactions at the surface of platinum at a pH range of 0 to 14. A successful project would provide insight into the reaction mechanisms on platinum, improve catalyst design principles, and extend the FE characterisation technique to alkaline solutions.

Supervisors: Signe Vahur, Rünno Lõhmus, Hilkka Hiiop

The knowledge about the composition of the materials used for cultural heritage objects is important for history (elucidating the origin, age, authenticity of the artefact) and conservation (for choosing suitable conservation materials). However, these materials have a very complex composition and analysis of organic materials, in terms of instrumentation, is challenging. In addition, non-destructive and direct surface analysis methods are preferred.
The main objective of this PhD project is the development of a laser-based pen probe mass spectrometric (MS) system that enables non-destructive analysis directly on the sample surface or surface ablation, allowing to remove the deposited material from the surface and its MS analysis. This PhD project is directly related to the comprehensive interdisciplinary PRG project (S. Vahur’s PRG1198 project) that engages chemists, physicists, engineers, conservation scientists, and archaeologists. Within this project, the possibilities of the analysis of cultural heritage objects will be expanded, and the final outcome will be a unique laser-based MS device.

Supervisors: Eneli Härk, Enn Lust

Sodium-ion batteries (NIBs) are potential candidates for stationary energy storage. Compared with lithium-ion batteries, NIBs can comprise cheaper and more abundant materials. Hard carbon is one of the most favourable anode materials for commercial NIBs. However, the process of the Na storage in hard carbons and the relation between the structure of hard carbon and the respective electrochemistry is still debated. This project aims to investigate the Na ion charge storage in various hard carbons by thoroughly characterizing the hard carbon powders (ex situ) and corresponding electrodes and conducting in situ and operando experiments with wide-angle and/or small-angle neutron and x-ray scattering methods on the Na-ion battery half cells. The main objective is to establish how large could the power and energy density of the NIBs be if the structure of the hard carbon is optimal.

Supervisors: Marek Mooste, Kaido Tammeveski, Margus Marandi, Ivar Zekker

Currently, one of the limitations for the large-scale commercialisation of the fuel cells is the use of Pt-based oxygen reduction reaction (ORR) catalysts. Therefore, an alternative non-precious metal catalyst material is needed for the replacement of Pt-based cathodes. Recently, the use of conducting polymers for the preparation of platinum group metal-free materials for fuel cell applications has grown due the attractive properties of the conducting polymer-derived ORR catalysts (CPDC) such as low-cost and facile preparation routes. For tuning the properties of CPDC materials towards higher ORR activity, doping with heteroatoms (e.g. S, N, P, B) and transition metals (e.g. Fe, Co, Mn) can be performed. The aim of the present PhD study is to prepare the sulphur and nitrogen codoped catalysts using the pyrolysed polymer blends for this application. To study the influence of the morphology of polymer materials on the ORR activity, chemical and electrochemical polymerisation methods will be both used. The electrochemical polymerisation is generally preferred for better control over the polymer properties. In present work, the environmentally greener solvents (e.g. water) will be preferred as the polymerisation medium. In addition, the inclusion of transition metals into the polymer structure will be employed to prepare highly active M-N4 type ORR-active centres into the pyrolysed polymer material. A thorough electrochemical and physical characterisation of the pyrolysed materials will be carried out together with the microbial and polymer electrolyte fuel cell testing.

Supervisors: Geven Piir, Uko Maran

Pesticides are controversial class of chemicals because they are deliberately released into environment to eliminate some specific living organisms. Followed by a quick decomposition into harmless chemicals. Unfortunately, many pesticides persist in environment for a long time, being deadly to other nontarget species. Unfortunately, it is not possible to test the harmfulness of pesticides in all species, which makes it difficult to assess the environmental risks they pose. Therefore, it is recommended that mathematical model predictions be used to provide preliminary risk assessment. These mathematical models are quantitative structure-activity relationships (QSAR) that link together information on chemical structure and experimental biological activity. These relationships can be represented as simple multi-linear regression as well as complex models based on machine learning (ML) and artificial intelligence (AI) algorithms. The aim of the proposed doctoral thesis is to systematically analyse and improve QSAR methods for assessing the environmental risks of pesticides based on a multi-step and/or multi-label model approach. The result of the dissertation is a broader understanding of the relationships between chemical structure and the properties studied, reliable QSAR models for the environmental risk assessment of pesticides, and innovative methodological solutions in the field of computational toxicology.

Supervisor: Ivo Leito

The recently introduced unified pH (pHabs) concept removes one of the main drawbacks of the conventional pH (IUPAC pH) – separate, non-comparable pH scales in different solvents. The main advantage of the pHabs scale is that pH values measured in any solvent become directly mutually comparable (in terms of the thermodynamic activity of the solvated proton). Because of this, pHabs might revolutionize the way we measure and interpret pH in non-aqueous and mixed solvents, including solvents with low polarity. This concept has been for the first time experimentally realized at UT and up to now pHabs measurements are done only in very few places in the world. Measuring acidity as pHabs is potentially highly advantageous in catalysis, liquid chromatography, sustainable energetics, rationalization of acid-base processes, etc. The PhD project aims at advancing the measurement methods, especially in low-polarity solvents, and demonstrating the usefulness of the concept on exemplary applications.

Supervisors: Ave Sarapuu, Kaido Tammeveski

Low-temperature fuel cells and metal-air batteries are efficient and sustainable energy conversion devices, but their commercialisation requires development of inexpensive non-precious metal catalysts for electrochemical oxygen reduction reaction (ORR) and oxygen evolution reaction (OER). Transition metal-containing nitrogen-doped carbon nanocatalysts are active for ORR and can be prepared by pyrolysis of organic precursors in the presence of nitrogen and transition metal sources. By including OER-active centres, e.g. transition metal oxides, sulphides or carbides, bifunctional ORR/OER catalysts can be prepared, which can be used in metal-air batteries. One of the most viable and cost-effective options is to employ abundantly available and renewable biomass sources as organic precursors for ORR and OER catalysts.

The aim of the project is to develop methods for preparing active nanocarbon-based non-precious metal catalysts for ORR and OER from biomass. The catalysts will be synthesised by pyrolysis of biomass, cheap nitrogen precursor (urea, dicyandiamide, etc.) and simple metal (Fe, Co, Ni, etc.) salts. The porous structure of the catalyst will be tailored for efficient O2 electrocatalysis and mass transfer, using template-based methods. The nature and ratio of the precursors and pyrolysis conditions will be optimised. The electrocatalytic activity of the synthesised materials will be evaluated and short-time stability tests will be carried out. The best performing catalysts will be extensively characterised by physicochemical methods, tested at anion exchange membrane fuel cell conditions and in metal-air batteries.

Supervisors: Gunnar Nurk, Kuno Kooser

Aim of thesis is development of redox stable mixed ionic-electronic conductive hydrogen electrodes for reversible solid oxide fuel cell. Deployment of this type of electrodes enable longer lifetime and simpler design of device and should decrease the system price because of that. More specific goal of this study is to clarify the dependence of electrochemical characteristics on chemical composition of electrode surface where surface chemistry is dynamically dependent on oxygen partial pressure, temperature, and electrode potential. This information is useful for designing optimal electrode compositions. For this purpose simultaneous chemical and electrochemical characterization of electrode surfaces will be carried out at conditions very similar to real operating electrode. To perform these experiments novel operando setup has been developed for impedance- (IS) and electron spectroscopic (ES) characterization. Three different semiconductive materials will be studied: La0.4Sr0.6TiO3-d is n-type semiconductor and has the best conductivity at very low oxygen partial pressures (~10-25 bar), La0.75Sr0.25Cr0.5Mn0.5O3-d is p-type semiconductor and performs well at moderate and high oxygen partial pressures and the conductivity mechanism of Sr2Fe1.5Mo0.5O6-d is dependent on oxygen partial pressure. Similarly collected comparative IS-ES dataset is very valuable for interpretation of electrochemical data of MIEC electrodes in general.

Supervisors: Koit Herodes, Asko Laaniste, Ivo Leito

Goal of the project is to develop a modern approach to analytical method validation with proof of concept software. It will rely upon and further develop the validation software ValChrom ( The outcome will be a published concept to automate validation process using the combined interdisciplinary strength of chemistry, software development and complex statistical approaches (design of experiments). PhD candidate will pull together these resources and systematize a modern approach on how to bring the validation of analytical methods out from the currently largely error-prone and cumbersome spreadsheet based software approach into automated and statistically sound approach. This will assume a high level of interdisciplinary knowledge handling and also familiarity with software development.

Supervisors: Heiki Erikson, Kaido Tammeveski

Energy demand grows every year and price grows along with it, this is further amplified by shut down of traditional power stations. This means that environmentally friendly energy conversion devices must be developed and low-temperature fuel cells, both with proton exchange membrane and anion exchange membrane, have proven to be suitable alternatives. Unfortunately, their wide spread use has been hindered by the amount and price of platinum, which is needed for efficient oxygen reduction on the cathode of a fuel cell. Additional issue is the long-term stability of the platinum catalysts. The aim of this project is to develop novel platinum-based catalysts with emphasis on stability and substrate effects. Namely, several titanium compounds as support could improve the long-term stability of catalyst and lower the amount of Pt in the material. This is supported by the fact that there are several studies with TiO2 that improves the longevity of the fuel cell cathode catalyst.

The focus will be on fundamental approach to investigate the oxygen reduction reaction on thin-film composite coatings in the beginning of the project. The second part of the project will deal with preparation of the catalysts supported on high surface area carbon that could be utilized in fuel cells.

Computer Engineering

Supervisors: Gholamreza Anbarjafari and Chagri Ozchinar

Small and tiny object recognition is a challenge that has become more important due to rapid development in AI based security tool development. In this PhD we are focusing on recognition of tiny objects in 2D and 3D using single shot learning. We will investigate Multiview analysis for 3D objects to enable late fusion in the recognition pipeline. 3D objects will be presented in pointcloud form which will be produced through the project SilentBorder. The developed deep neural networks will be subject to network optimization to assure the developed solutions can be used in the real-time scenarios.

Supervisor: Arun Kumar Singh

Motion planning and control algorithms are essential building blocks of any robotic application. They are often formulated as an optimization problem as it allows for encoding higher level robot behaviours through appropriately defined cost and constraint functions. A fundamental challenge in robotics is to make optimization-based motion planning reliable and real-time. The reliability issue stems from the fact these optimization problems have multiple solutions (minima) and finding the best among these is computationally challenging, ,especially under real-time constraints. The proposed PhD thesis will achieve major breakthroughs in solving non-convex optimization problems encountered in robotics. The application domain of the research will be diverse ranging from unmanned aerial systems to autonomous driving and manipulation.

Superors: Karl Kruusamäe, Arun Kumarvis

A human-robot collaboration is envisioned to leverage the strengths of humans and machines in the manufacturing industry and beyond. In a collaborative work setting, it is impractical to assume that the human will use conventional controls and displays to interact with the robot as it interrupts their individual role in the collaborative process. Instead, human perception and performance is enhanced by integrating the worker into a sociotechnical cyber-physical system, a concept that is often referred to as Operator 4.0. Augmented reality (AR) is considered as the enabling technology that addresses both human perception and performance in Operator 4.0. AR headsets combined with gesture- and verbal-based control provide intuitive, yet personalized means of collaboration for humans and robots. The objective of this project is to design, validate, and publish AR-based humanrobot interaction for manipulator-assisted collaborative assembly and sharedcontrol for a mobile robot (e.g. payload transportation). The project will set out to collect and publish relevant human-robot interaction datasets and design UX validation frameworks which are re-usable and enable cross-platform benchmarking.

Supervisor: Alan Henry Tkaczyk

This PhD project will develop Dynamic Systems Models (DSM) to assess environmental performance of recycling processes for industrial materials. Recycling of industrial residues (e.g., materials, by-products, and wastes) is an economically-sound option to meet sustainability goals, but associated environmental impacts must be quantified, to prevent the creation of a new problem while solving the first problem. For example, re-use of industrial residues containing heavy metals, radioactive residues, or toxic compounds can have unintended environmental or safety consequences, if the full life cycle of the target applications is not comprehensively assessed. A challenge and opportunity in these recycling efforts is to ensure economic viability and compliance with environmental standards. This thesis proposes a novel DSM methodology which could address such concerns in mining, recycling, and chemical processing industries. The aluminium industry annually produces 150 million tons of Bauxite Residue (BR or “red mud”) as an industrial by-product. At the same time, BR has high alkalinity, is a waste management challenge, and falls under the legislation for Naturally Occurring Radioactive Material (NORM). Instead of being landfilled, BR could be analyzed, treated, and used as a recycled input material for sustainable cement production. The research objective is to devise a novel systems modeling methodology to assess the environmental performance of recycled industrial materials in real-world scenarios including time-dependent variations in input feeds and changing guidelines. To ensure that the new BR recycling technologies deliver the maximum environmental benefit and minimize the risk of creating secondary harmful effects, evaluation of their potential impact using DSM will be completed. The work will integrate two research fields (Life Cycle Analysis & NORM) during the development of the DSM methodology. Implementation of such methodology is an ambitious and challenging task to fulfill, having important impacts in terms of interdisciplinary academic publications and practical outcomes for industry and society. A Euronews video and more information is available at: and

Supervisors: Arun Kumar Singh

It is well known that human beings walk or move their arm joints optimally. In other words, human motions follow the framework for optimal control. Moreover, it is fascinating how they can transfer the optimal movements to novel tasks, environments, and contexts. The proposed Ph.D. project will develop algorithmic and technological solutions towards imparting the same flexibility to robots. To this end, the project will bring together diverse concepts from sensitivity analysis of optimization problems, deep generative models, and sequence to sequence prediction from Natural Language Prediction. The project results will make robots resilient to the environment and task changes and potentially open-up new application areas.

Supervisors: Karl Kruusamäe, Leo Aleksander Siiman

Text of summary (ca ½ pages) There is an intriguing dilemma in the technologically-improved modern world: using technology gets simpler and simpler while the technology itself is growing more and more complex. This complexity poses a continuous challenge to educating engineers required to create and troubleshoot advanced technological solutions. Research in higher education STEM courses indicates that much of the variation in student outcomes can be explained by variations in students’ preparation (also referred to as prior knowledge). Finding effective ways to differentiate instruction based on the incoming preparation levels of learners is urgently needed to increase retention rates and minimize dropout across a variety of STEM disciplines. This PhD project addresses two main research questions: (a) Which teaching and learning practices are effective at differentiating instruction for learners with varying levels of prior preparation? (b) To what extent can a university robotics course that applies principles of differentiated instruction reduce learning gaps among students’ with varying incoming preparation?

Computer Science

Supervisors: Eduard Barbu and Raul Vicente Zafra

As machines take more and more decisions that once were handled by human experts, there is an increasing need to understand the process of machine decision-making. Only the interpretable machine learning models can be trusted, are safe, and contestable; that is, people can appeal to machine decisions.

This thesis aims to improve the state-of-the-art AI explainability by injecting background knowledge into the explanation engine for a machine learning trained model.

The student will test the solution on two use-cases part of the European project Trust-AI. The first use-case intends to explain the forecasted electric energy demand and cost for a particular area. The second use-case aims to develop and explain a machine learning model that predicts tumor onset and tumor growth for human patients. For this purpose, the structure of the explanation will be formalized using insights from cognitive science and psychology. Based on the formalization, background knowledge will be extracted from domain-specific corpora. Together with an enhanced SHAP interpretation framework, this knowledge will allow answering pertinent questions about the use-cases under study.

Supervisors: Jaan Aru, Kadi Tulver

Thanks to advances in technology, the use of various devices has greatly increased in various fields as both their accessibility and portability have improved. Using such technologies in health care can greatly increase the availability of treatment which is especially critical in the field of mental health. One such technology is virtual reality (VR), which has shown potential as a complementary or even an alternative method for treating various patient groups, e.g., those with addictive  disorders. We hypothesize that certain VR environments can alleviate addictive disorders. This hypothesis is based on the group’s finding where a specifically developed VR had a significant positive effect on depressed participants’ emotional state and psychological insights. In order to test this hypothesis we will conduct a randomized clinical trial comparing the effects of our novel VR experience and ordinary VR. We will collaborate with clinical psychologists and health institutions. The goal of this doctoral thesis is to help in developing a new method for the treatment of addictive disorders.

Supervisor: Raivo Kolde

Real-world medical data is a treasure trove of information for research. However, the information is not easy to access as much of it is kept as free text notes. The bottleneck for information extraction from such data is the annotation and labelling of the datasets. Modern pre-trained deep learning based language models can be relatively accurate even on small training sets, creating the opportunity to develop human-in-the-loop type systems for data labelling. In such system, an expert reviews the results of a machine learning model through several iterations of the training, thus, guiding the model towards more accurate results. A well-designed, could eliminate the need of a text mining or machine learning expert in the workflow, speeding up the process considerably. The goal of the PhD project is to design and implement such system for medical texts. The machine learning component will be based on BERT like language models, pre-trained on massive amounts of medical texts. However, the real focus of the project is building and validating the labelling and training infrastructure, specifically geared towards medical texts. It is also important to build a backend that allows to store the learned models, generated test sets, and pipeline results, while facilitating further refinement of the models and their use in other datasets. The resulting software will be utilized in ongoing research collaborations, where the goal is to harmonize cancer datasets from variety of source across Europe, to enable machine learning for clinical decision support.

Supervisor: Sulev Reisberg

Although health records are mostly in electronic format nowadays, it is still challenging to effectively extract information from these records and ensure the high quality of the data. This makes further analysis of the data difficult and time-consuming. For instance, in Estonia, health information is collected to central e-health database Tervise Infosüsteem still as CDA HL7 clinical documents which are relatively loosely structured. Even in the structured fields, the data format is rarely checked, leading to tens of terabytes of data which is hardly usable for any epidemiological or statistical analysis. Although some efforts are made to improve this situation, the solutions have not been effective. The aim of this doctoral thesis is to address this issue and provide solutions for better health data extraction and management.

It develops methods and provides implementations of these methods to extract information from CDA HL7 document, clean the data and map to common health standards/classifiers.

Supervisor: Meelis Kull

Modern deep neural networks are well known to be over-confident due to overfitting on the training data, under-estimating uncertainty in their predictions. It is commonly addressed by post-hoc calibration methods that calibrate the confidence of neural networks using holdout validation data. This approach succeeds on future data from the same distribution as validation data, but fails on out-of-distribution (OoD) data. For example, models trained to classify items of clothing can confidently classify an image of a hand-written digit into a particular class of clothing items. OoD over-confidence poses a huge risk for autonomous vehicles, since their decision making systems are relying on the uncertainty reported by the neural networks within the perception system. Our objective is to develop methods that recognise different contexts from the data and perform contextual uncertainty quantification and calibration, taking into account the amount of experience in various contexts in the training process. We aim to align our methods with the applications in autonomous driving, a field where the safety of subsystems involving machine learning is particularly important.

Supervisors: Kuldar Taveter, Eva Maria Navarro-López

This proposal for a doctoral research project is targeted at designing and implementing emotion-aware autonomous agents. The difference from the research work done so far is that the architecture will be based on the theory of predictive probabilistic brain, which is considered as the new state-of-the-art in neuroscience. The theory of predictive probabilistic brain can fundamentally change the way software agents and chatbots are designed and implemented from traditional symbolic architectures towards probabilistic architectures. However, computational implementation of predictive probabilistic processing is still faced with problems of intractability. Against this background, this doctoral project aims to work out ways how to manage with the problems of intractability of predictive probabilistic processing by using, for example, corpora of predefined emotion-related words and expressions, and work out appropriate algorithms of machine learning for processing and expressing emotions. Other goals of the project are to find out how to represent the knowledge required for reasoning by emotion-aware agents in a generic and compositional way and how to support the design of emotion-aware agents by the appropriate methods of requirements engineering and software design. To answer the research questions, we will apply the research method of action design research that is aimed at finding a solution to the problem and embodying the solution in the artefacts – emotion-aware software agents – and evaluating the artefacts iteratively with end users. The resulting emotion-aware software agents will be targeted at different areas, including interactions with older adults to overcome loneliness, which is a topic of the ongoing H2020 project of the European Union “Pilots for Healthy and Active Ageing”, which involves UT.

Supervisor: Arnis Paršovs

Interoperability is the backbone of modern society and e-Government. Interoperability is required between systems which range from core governmental registries to the smallest IoT gadgets. Federation of systems impose security, privacy and technological challenges which must be investigated, resolved and communicated to interested parties.

The Estonian-invented X-Road framework has been in use for 20 years. The last generation of X-Road was analyzed and implemented in 2012-2014 and since then has received only minor updates and face-lifts. The new paradigms for system integration and technology require deep analysis of the existing shortcomings to find possible solutions.

The main focus of the research is the Estonian use case of the X-Road model, but during the research, comparisons with similar solutions abroad would also be performed.

Supervisors: Riccardo Tommasini, Kalmer Apinis

Dealing with large amounts of data is one of the biggest challenges that permeate computer science, with much academic and commercial interest. When faced with this dilemma, the first step is to try to reduce the amount of data.

A novel reduction method, materialization, is thoroughly used in fields with heavy usage of ontologies, such as medicine and avionics. Using materialization, one physically stores only a subset of the whole data and derives the rest of it ad hoc.

However, a big problem with current materialization methods is scalability. These methods can not handle enough data to be useful in practice, or the data has to be static -- which is often an unrealistic scenario.

Nevertheless, differential dataflow[1], a bleeding-edge model of distributed computation, is a promising approach to tackle this exact problem. The thesis revolves around building a materialization system on top of it, ultimately validating its usefulness and suitability in a wide range of scenarios.

Environmental Technology

Restoring degraded peat soils is an attractive, but largely untested climate change mitigation approach. Drained peat soils used for agriculture or for peat extraction are often large greenhouse gas sources (GHG). Restoring subsided peat soils to managed, impounded wetlands can turn these sources into carbon (C) sinks. However, at present, the amount of scientific information available to guide such restoration decisions and assess the impact of these actions is still sparse and restoration outcome can be low C uptake and high methane (CH4) and nitrous oxide (N2O) emissions. Therefore, the overarching objective of this study is to provide an experimental and theoretical understanding how to restore wetlands with minimized CH4 and N2O emissions and maximized C uptake. The aim of this is PhD project is to study two restoration strategies in temperate climate and to understand, which methods provides fastest C sequestration: a) using fast growing macrophytes and, b) Sphagnum mosses. Both sites will be continuously monitored for CO2 and CH4 flux with eddy covariance (EC) technique and with chamber technique to understand the heterogeneity of fluxes. A thorough understanding of CO2, CH4 and N2O flux drivers is essential for these restoration projects because the magnitude of the emission of these gases can determine if these restored systems are GHG sources or sinks. The research will be carried out in three restored wetland ecosystems in Estonia but with strong collaboration with international partners in Europe, Korea, and USA.

Supervisor: Kuno Kasak

Restoring degraded peat soils is an attractive, but largely untested climate change mitigation approach, especially in large extensive abandoned peat-extraction areas. Drained peat soils used for agriculture or for peat extraction are often large greenhouse gas sources, especially carbon dioxide (CO2) and nitrous oxide (N2O). Restoring subsided peat soils to managed, impounded wetlands can turn these sources in to CO2 sinks but at the same time the emission of methane (CH4) can increase several orders of magnitude. At present, the amount of scientific information available to guide such large-scale restoration decisions and assess the impact of these actions is still sparse and restoration outcome can be low carbon uptake and high CH4 emissions. Therefore, the overarching objective of this PhD project is to provide an experimental and theoretical understanding how to restore wetlands with minimized CH4 and N2O emissions and maximized carbon uptake. To develop restoration recommendations, we will analyze multiple drivers for greenhouse gas emissions including site specific hydrology, legacy effects, soil chemistry, soil microbiology, vegetation development and dynamics to understand the hot spots and hot moments of GHG fluxes in the restored wetlands. The PhD students will work at two recently restored wetlands ecosystems. The PhD students will use eddy covariance tower data with chamber measurements to understand how different restoration strategies (e.g. fast growing emergent macrophytes, Sphagnum mosses) affect CO2 sequestration efficiency and the emission of CH4 and N2O. The specific aim for this study is: 1) To detect hot spots and hot moments of CH4 and N2O emission in the restored wetlands and to understand the regulating mechanisms; 2) To analyze how the vegetation development varies in the wetland and how it determines carbon sequestration efficiency and emission of CH4 and N2O, and 3) to carry out field scale experiments to control the wetlands GHG emissions.


Supervisors: Elin Org and Raivo Kolde

The gut microbiome is a complex and metabolically active community that directly influences various host phenotypes and in recent years our knowledge about the associations between commensal microorganism and host physiological processes has advanced rapidly. Diseases being associated with alterations in the microbiome composition range from type 2 diabetes, IBD and colorectal cancer to autism and Parkinson’s disease. Due to the seen associations, microbiome based risk scores have become a topic of interest for identifying people at risk. However, the host-gut microbiome interactions, including interactions between host genetics, metabolic activity and gut microbiome, remain largely unknown. Understanding the complex interactions and taking them into account during the risk modelling could lead to substantial increase in prediction accuracy and advancement of personalized medicine.

The goal of the current project is to understand the role of microbiome in complex diseases through investigating the interactions between human gut microbiome and various –omics data, including host genetics and metabolic activity, and to evaluate the impact of multiomics integration in risk models for complex diseases. Estonian Biobank cohort data will be used for the project, with gut microbiome data and plasma NMR profiles available for more than 2500 individuals in addition to genotyping and extensive baseline and follow-up data. In addition, the project will use data collected from a project investigating the role of microbes in the development of colon cancer.

Supervisors: Reedik Mägi, Maris Laan, Margus Punab, Triin Laisk

Male infertility is a common, complex disease, affecting ~7-10% of men, and manifesting in diverse phenotypes ranging from morphological and functional abnormalities in sperm to severe spermatogenic impairment, in which few or no sperm are produced by the testis. An abundance of epidemiological data indicates that male infertility is often not an isolated condition, and it is not only a concern related to failed fatherhood. The accumulated evidence provides solid understanding that male infertility and overall health are interconnected.

This interdisciplinary project is expected to improve the understanding on shared genetic determinants, etiologies and involved molecular mechanisms leading to impaired reproductive and overall health in men. The project data will link genetic causes of male infertility and other reproductive traits and diseases to overall health measures and chronic disease risks across lifetime.

Supervisors: Vasili Pankratov, Luca Pagani, Francesco Montinaro

Isolation, migration and local adaptation lead to the emergence of local genetic structure (allele frequency differences) in the Estonian population. Understanding such genetic structure is essential not only for revealing the population's past but also for medical and epidemiological practises.

Recent studies found that, even when strictly accounting for the genetic structure of a population, the genetic liability for some phenotypes can still show a significant correlation with geography.  A possible interpretation for this feature is that very recent migration might have, at least in part, influenced the allele frequency of genetic variants involved in the manifestation of certain phenotypes. As an example, the Polygenic Risk score for the BMI in people born and living in deprived areas in the UK is significantly lower than people that have born in the same area but subsequently moved. 

Another, deeper layer of variance is represented by uneven contribution of ancient populations to the various Estonian counties. The impact of ancient genetics to the phenotypic landscape of modern Europeans has been recently investigated in Estonian biobank and will constitute essential background information, which will also be feedbacked to the Biobank participants  during the course of this project.

This PhD project aims at a) investigating the interplay between recent migrations and distribution of Polygenic Risk Scores for various phenotypes in the Estonian population and b) developing and testing a strategy of communicating personalised feedback on ancestral populations’ contribution to particular phenotypic traits of a given individual. This will not only deepen our understanding of the origin and consequences of genetic structure in Estonia but might also have practical value for understanding local differences in prevalence of various heritable conditions. Providing a novel type of genetic feedback for the biobank participants will help to engage both current and novel biobank participants, potentially facilitating recruitment for future research projects.

Supervisors: Uku Vainik, Tõnu Esko

Typical tendencies of thinking, feeling, and behaving can be summarised by personality and cognitive traits and measured with tests. These behavioural traits associate with life outcomes, such as health and well-being. However, causal directions of these associations are largely unknown, as randomised controlled trials are resource-intensive and often ethically impossible. We propose using a genomic causal inference method – Mendelian randomisation. This method capitalises on natural randomisation of genetic variants – genetic lottery – that cause differences in behavioural traits. The result will be an atlas of plausibly causal associations between behaviour and life outcomes. The compiled dataset will also be a valuable public resource for any researcher interested in the intersection of behaviour, genetics, and health research for many years to come.

Supervisors: Tõnu Esko, Erik Abner

Estonian biobank (EstBB) was set up as a population-based biobank, with the intention to spearhead the implementation of personalized medicine in Europe. To date, over 205 000 consented participants from all demographic distributions within Estonia (around 20% of the whole adult population) have voluntarily provided biological materials to EstBB. All the blood samples from the biobank participants have undergone DNA microarray-based genotyping and are periodically linked to national electronic health records.

The genetic data of the biobank participants has been imputed utilizing a local genomic reference panel, which allows for a considerably higher imputation quality than with other non-population specific reference panels. As such, reliable genetic data on thousands of missense, stop-gained, loss-of-function and other functionally relevant variants has recently become available. However, a large proportion of these rare genetic variants have only been poorly described. The electronic health records per biobank participant generally span for over a decade, allowing EstBB researchers to statistically associate these variants to common health traits and disorders. For example, an initial analysis of these rare variants has allowed us to identify a rare genetic variant, which is significantly associated with obesity and might affect the life quality of 1-in-115 Estonians.

The aim of this thesis is to run a broad-scale analysis on the rare genetic variants present among the Estonian population, with the objective to characterize their effect on common complex diseases. The results from the broad-scale analysis of rare variants will allow us to consider additional in-detail studies of relevant functional genetic variants. Furthermore, this thesis will also concentrate on thoroughly analyzing the biological and clinical aspects of the previously mentioned obesity causing variant.

Utilizing the already existing genetic and phenotypic data from Estonian biobank, this thesis will further our understandings of the molecular etiologies of complex diseases and assist in the implementation of personalized medicine in Estonia.

Supervisor: Kelli Lehto

Since there are no biomarkers or other biological tests for diagnosing psychiatric disorders, clinical diagnoses for the most common psychiatric disorders, such as depression, are based on a wide range of heterogeneous symptoms, which may drastically differ between patients. The large heterogeneity within and non-specificity across psychiatric diagnostic categories is complicating the precise diagnostic and treatment decision-making for clinicians. We are currently lacking both, the understanding of clinical subtypes within larger disorder categories, and the methods to identify those subgroups clinically, which would assist mental health specialists in making more accurate diagnostic and drug prescription decisions. The aim of this doctoral project is to jointly analyse genomic data, diagnostic and medication prescription information from the electronic health records and questionnaire-based self-reported symptom level data to identify psychiatric disorder subtypes corresponding to underlying biology and develop prediction models to enable improved disease stratification in the future. Genetically informed prediction models would be highly valuable novel tools for psychiatrists and other mental health specialist to assist improved clinical decision-making in psychiatry.

Supervisors: Neeme Tõnisson, Tarmo Annilo

The aim of the project is to understand genetic and environmental factors that modify phenotypic heterogeneity of monogenic disorders. It has been shown earlier that modifier factors, including polygenic background, can alter the phenotype of monogenic disorders such as cystic fibrosis and Stargardt disease. Our workgroup is studying several diseases (e.g. long QT syndrome, familial breast cancer, Wilson disease), where we will identify and estimate the functional impact of variants in disease-relevant genetic pathways using data from GWAS studies and whole-genome/exome sequencing efforts. We will also analyze metabolome, transcriptome, DNA methylation and environmental contributors to identify associations with phenotype severity. Using Electronic Health Record data, we will generate prioritized gradual disease trajectories which will be analyzed in the context of genetic and multi-omic modifier factors.


Supervisors: Evelyn Uuemaa, Alexander Kmoch

Agricultural production is one of the major forces of global environmental degradation and intensive agriculture drives nutrient losses to groundwater, streams, and rivers, which in consequence has become the main cause of eutrophication in waterways and coastal zones. The aim of the project is to develop spatio-temporal machine learning-based water quality models for national (Estonia), regional (Europe) and global level. In this project, the candidate will develop interpretable machine learning models that enable to identify the complex co-variances and the most important features that contribute to water quality in the streams. And building on that, the candidate will develop robust and reliable method to predict nutrient runoff at large scales and accordingly estimate the accumulated loads along the water pathways and pressures on coastal zones. To achieve the aims of the project, the candidate will use remote sensing and existing data products (climate, soil, land use etc.) to prepare spatially gridded time-series as input layers into a data cube. The data cube will be used to build the machine learning based model to explore and explain relationships of the key factors influencing nutrient concentrations and predict nutrient runoff at different scales and resolutions (from national to global level).

Supervisors: Kadi Kalm, Tiit Tammaru, Kadri Leetmaa

In recent years, the linkage between segregation and gentrification processes has received a lot of attention. However, there is no clear understanding of the relationship between these two and clarification is needed about the impacts of gentrification on contemporary segregation and migration patterns. The present PhD project contributes to this gap and questions whether and on what conditions do the displacement mechanisms of gentrification provide explanation for the social, age and ethnic homogenisation of post-socialist neighbourhoods, and how does urban displacement related to counter-urbanization processes. It aims to develop a fuller understanding of the relationship between gentrification, migration and segregation based on the case of Estonia and its capital city Tallinn. The capital of Estonia has experienced fast social and spatial change during last quarter of a century and as a context that can be described by superhomeownership it offers a great place to study the relationship between these two processes. Data analysis is based on the individual-level census databases (1989, 2000, 2011, 2021) and recent registry data (2015-2021).

Supervisors: Ivika Ostonen-Märtin, Priit Kupper, Marika Truu

Understanding and predicting how climate and land-use change affect ecosystems' structure and functioning is a crucial challenge of the 21st century. Environmental change (such as rising temperatures and humidity, frequent weather extremes, drainage of wetland forests, or increasing frequency of Ips typographys attacks in spruce forests) in northern latitudes have consequences for important forest ecosystem functions, such as plant biomass production and shifts in above- and belowground allocation patterns, which may feed-back to climate through shifts in the biogeochemical cycles in ecosystems. Qualitative or quantitative changes in rhizodeposition (root litter, metabolite exudation) will cause a shift in the structure and function of soil and rhizosphere microbiome responsible for stabilizing SOM. Plant roots and their associated microbial communities play a significant role in the rhizosphere and soil nutrient cycles that need to better understood from the point of view of the sustainable functioning of forests as well as of global C cycle in soils.

The overall goal of this doctoral project is to advance our understanding of the effects of changed water conditions (drainage, irrigation, higher relative air humidity), higher temperature, weather extremes (heatwave, flooding, drought) or higher frequency of insect’s attacks’ on trees growth and on the structural and functional adaptation of plant and soil microbiomes related nutrient cycle in the rhizosphere.

The novel tools for soil and rhizosphere solutes sampling (microdialysis) and data analysis approaches including machine learning will be applied to analyse and integrate nutrient fluxes in the rhizosphere (including interactions between plant and soil microbiome, plant traits and physiological processes and environmental conditions). Novel and advanced approaches in belowground studies help to bring the changes in plant root traits, soil microbiomes and rhizosphere C and nutrient cycles to the ecosystem level.


Supervisors: Leho Ainsaar, Aivo Lepland

Carbon and oxygen stable isotopic composition of marine carbonates has widely used as proxy for the palaeoenvironmental conditions. Secular trends and perturbations in carbon isotopic composition are mainly interpreted as temporal changes in the global carbon reservoir and principal processes of sequestration with implications to palaeoclimatology. However, studies on several recent carbonate environments have described and explained cases where carbon isotopic composition of carbonates differ along onshore-offshore profile. This has led to discussion if secular isotopic changes in geological sections have been caused or influenced by changes in depositional settings and/or early diagenetic environment. The oxygen isotopic composition of sedimentary material has been used to estimate the palaeotemperatures, but the preservation of this signal from secondary alteration is even more questionable than in case of carbon isotopes. The project is focussed on the analysis of isotopic composition of different sedimentary (bioclasts, mud) and secondary (calcite cements, dolomite) components of selected Baltic Palaeozoic carbonates. The intervals of Ordovician and Silurian palaeoenvironmental events with increased heavy carbon isotope concentration are selected for the target. To assess the alteration effects of cementation upon the bulk geochemical and isotopic signatures and to extract the robust primary signal we plan to undertake a dedicated study of individual carbonate generations using a variety of in situ techniques in Tartu and in partner universities.

Supervisors: Martin Liira, Päärn Paiste

The Baltic Sea is one of the most anthropogenically impacted seas globally. According to the Marine Strategy Framework Directive (MSRD), Member States must prioritize achieving/maintaining good environmental status in marine areas and preventing further deterioration. The seabed and sediments are an integral part of the marine environment. Most coastal water bodies are not still in a good environmental state in Estonia. At the same time, the development of offshore wind farms in Estonia has been set as a priority for the coming years. So far, data about Estonian seabed sediment geochemistry is limited and does not meet the requirements to assess the environmental status of those sediments and, therefore, estimate and mitigate the mobilisation of potentially environmentally hazardous components. Phosphorus is present in seabed sediments in several forms, but only part of it is mobile, i.e., it can be released back into the water, causing additional eutrophication.

Supervisor: Kalle Kirsimäe

The aim of the project is to study the redox conditions dynamics and trends in at the beginning of the Proterozoic period when the free dioxygen first appeared in the Earth’s atmosphere triggering the cascade of changes in Earth’ (bio-) geochemical cycles. The project focuses on sedimentary succession in the Paleoproterozoic Franceville Basin, Gabon, which holds arguably the best preserved sedimentary successions of the Paleoproterozoic with only mild metamorphic overprint.

Supervisors: Alar Rosentau and Tiit Hang

The changing climate, rising sea level, increasing storminess, and possibility for tipping points leading to cascading, dangerous scenarios, call for better understanding of the functioning of the Earth System in the past, present and future. While meteorological records are short, the tideless and uplifting coast in the central Eastern Baltic Sea is unique because the elevated coastal formations and sediments serve as archives for deciphering past sea-level change and climatic events. The proposed PhD project aims to reconstruct the Holocene (last 11700 years) changes in the relative sea-level and storminess in western Estonia, based on analyses of the prograding coastal landforms and sediments. During the PhD project, new geomorphological and sedimentological evidences will be collected and analyzed from the selected coastal sections from western Estonia (including West Estonian Archipelago) to reconstruct palaeo sea-level and storminess.

Supervisors: Kaarel Mänd and Anthony Chappaz

The aim of the project is to study the speciation of vanadium and molybdenum in Baltoscandian metalliferous black shales (Alum Shale) in order to resolve spatial and temporal redox variations on the Cambrian-Ordovician boundary in the Baltic basin. The Alum Shale is present from Denmark and Sweden to Estonia and northwestern Russia, but the paleoenvironmental setting where it formed is as of yet unresolved, as is the history of redox variations on the biologically tumultuous Cambrian-Ordovician boundary in general.

Materials Science

Supervisors: Alvo Aabloo, Tarmo Tamm

Smart textiles have been a research topic and a possible technology for decades. The functionality of smart textiles can vary, one of the many possibilities being a chemically and mechanically active network of transducers, an exoskeleton, or a compatible biomedical device. Exoskeletons can be supporting structures to provide better capabilities while also being a network of electrochemical sensors Conductive polymers have been used directly as electrodes in many applications. However, in biomedical applications, the penetration of oxygen through the comfort of the material is very important, which is technically called breathability. To solve these problems, conductive polymers are used as the fiber form and driven. Depending on the knitting pattern, the flexibility of the material can be further improved, which contributes to better breathability. Because materials are sensitive to synthetic conditions, it is difficult to obtain reproducible results. The production of impurities (3D printing) incorporating the technology of the Fourth Industrial Revolution enables the synthesis in a highly complex structure with high resolution, rapid and economical prototyping and reduction of the human factor.

Supervisors: Andreas Kyritsakis, Veronika Zadin, Flyura Djurabekova

Vacuum arcs, i.e. electric discharges appearing in vacuum (also known as breakdowns), are a major limiting factor for various applications such as particle accelerators, fusion reactors, vacuum interrupters, X-ray sources, and space technology. However, the physical mechanism underlying the very initiation of the phenomenon still remains unclear. Recent experimental evidence indicates that the distribution of electromagnetic power is actually the main limiting factor of the arc initiation, instead of applied electric field and the cathode material as previously assumed. This project aims to understand the physics underlying the power supply limitations on the vacuum breakdown initiation by developing computational models that predict its behaviour. A direct comparison with experimental results shall result in the exact determination of the desired design characteristics for structures that suffer from vacuum breakdown.

Supervisors: Veronika Zadin, Sergei Vlassov

The behavior of nanostructures at elevated temperatures can differ drastically from the macroscopic situation. Nanostructures melt at a significantly lower temperature compared to bulk, and the melting temperature depends on size and treatment time. Surface energy minimization driven by thermally activated diffusion and Rayleigh instability can lead to drastic changes in morphology. For instance, fragmentation of Au and Ag nanowires appear at temperatures as low as few hundreds degrees Celsius and can hinder the functioning on nanowires-based devices like e.g. transparent conductive films. On the other hand, fragmentation of NWs into ordered arrays of nanoparticles can be utilized in various applications including biosensors and photodetectors. In the planned work, heat-induced modifications in the nanostructures will be investigated in realtime in-situ inside SEM using dedicated heating stage enabling fine control over the process and providing essential information on the dynamics. The effect of various covalent coatings on metallic NWs for the ability to tolerate elevated temperatures will be tested. Moreover, thermallydriven evolution of fine features (protrusions, outgrowths, tips etc) on the surfaces of bulk metal samples will be studied in collaboration with CERN. Such features are of high importance e.g. in the field of particle accelerators where electric-field-induced surfaces outgrowths are considered as potential cause of vacuum breakdowns.

Mathematical Sciences

Supervisors: Rainis Haller and Andre Ostrak

It is remarkable that there exist Banach spaces where every slice of its unit ball has diameter equal to two. Such Banach spaces have the diameter two property. In recent years, the family of different diameter two properties (usually natural strengthenings of the above one) has grown rapidly and has been extensively studied.

The proposed doctoral project aims to advance our knowledge on geometric structure of Lipschitz functions spaces and their canonical preduals, the Lipschitz-free spaces. Geometry of Lipschitz spaces has become a very active research subject. Every Lipschitz function between metric spaces admits a canonical linear extension between the corresponding Lipschitz-free spaces. This fundamental property makes Lipschitz-free spaces an efficient tool to study Lipschitz functions, which appear in many contexts.

Our primary objective in this project is to understand whether diameter two properties are indeed different in Lipschitz spaces. We conjecture that for the free spaces they are primarily the same but for functions spaces primarily different. The secondary goal is to characterise diameter two properties in terms of the geometry of the underlying metric space.

Supervisors: Rainis Haller and Märt Põldvere

Following Acosta et al., a pair (X,Y) of Banach spaces is said to have the Bishop–Phelps–Bollobás property (in brief, the BPhBP), if, for every real number ε>0, there are real numbers η(ε)>0 and β(ε)>0 with β(ε)→0 if ε→0+ such that given a bounded linear operator T:X→Y with ∥T∥=1 and an element x∈SX satisfying ∥Tx∥>1-η(ε), there are a bounded linear operator T0:X→Y and an element x0SX satisfying T0x0∥=1, ∥T-T0∥<ε, and ∥x-x0∥<β(ε).

During the last decade, the BPhBP for pairs of classical Banach spaces has been extensively studied. However, the constants η(ε) and β(ε) that can be found in literature very often seem to be not the optimal ones. This project aims at improving these existing known constants in many cases by giving new proofs to Bishop–Phelps–Bollobás-type theorems in literature.

Supervisor: Karin Täht

In a world which is constantly changing, learning mathematics is vital. Unfortunately, many students have mathematics anxiety which deter them from obtaining mathematical and other STEM subjects knowledge. Mathematics anxiety can affect students in their everyday life, such as in maths class and everyday situations as well both in school/maths class and in everyday life. The more anxious a person is about mathematics, the less maths subjects they will choose to take in school. Having mathematics anxiety can even prevent students from choosing their careers in STEM subjects. This means it is vital to research about mathematics anxiety and find ways to reduce or prevent it. This thesis will give an overview about mathematics anxiety and its relations to other learning attitudes and achievement. The purpose of the thesis is to work out intervention methods which can help teachers to reduce students’ negative emotions while learning mathematics.

Supervisor: Kristi Kuljus, co-supervisor Bo Ranneby

In statistical modelling it is essential to find a model that is in some sense most suitable for describing reality. Most model selection methods are related to some information criterion. Information criteria such as Akaike or Bayes select the best model among the candidates but they cannot measure whether the assigned model class is suitable. Maximum spacing method (MSP) which is originally a parameter estimation method for continuous distributions, enables to decide about suitability of the assigned model class for given data. The MSP method can be studied for different information measures which is essential in model validation context. One aim of the project is to study properties of MSP estimators for different information measures in the case of multivariate independent and identically distributed (iid) observations.  The second goal is to study extensions of MSP method to more general models with dependent observations, and explore its possibilities as a model validation tool for model classes of different complexity.

Molecular Biosciences

Supervisors: Maia Kivisaar, Heili Ilves

The aim of the proposed doctoral project is to investigate mutagenic processes and tolerance of bacteria to harmful chemicals in connection with environmental signalling. We will elucidate mechanisms of survival and evolution of bacteria under stressful conditions by using Pseudomonas putida as a model organism. The metabolic versatility of P. putida makes this organism attractive for biotechnological applications such as biodegradation of environmental pollutants and synthesis of added-value chemicals (biocatalysis). One of the specific tasks of the proposed doctoral project is to investigate mechanisms of increased mutation frequency in bacteria lacking two-component regulatory GacS/A system, which is known to regulate a large number of genes as a response to environmental stimuli. The second task is focused on elucidation of mechanisms affecting P. putida tolerance to harmful chemicals. The expected results could be important to understand basic mechanisms of evolution of bacteria and provide knowledge to be used in biotechnological applications.

Supervisor: Riho Teras

The regulation of the biofilm of the environmental bacterium Pseudomonas putida by extracellular peptides and the regulation of the expression of genes important for biofilm formation such as lapA-lapB and gacS-PP1651 by the exclone and the global regulator Fis is investigated.

The innovativeness and importance of the study is the new perspective for biofilm regulation that has not been addressed before. Namely, the study will assess the dependence of P. putida on environmental factors and the expression of critical genes by excludon. Excludon is the regulation of the expression of oppositely transcribed genes at the 5 'end of gene mRNAs that are complementary and act as asRNAs. The expression of the two gene pairs lapA-lapB and gacS-PP1651 could be regulated by excludone. Since we have found Fis binding sequences in both regions, it is aimed to determine how Fis affects gene expression based on excludon or RNAP competition.

Supervisors: Priit Väljamäe, Silja Kuusk

Lignocellulose represents a huge reservoir of renewable carbon. Its enzymes-aided valorization provides a green and sustainable alternative to the traditional, petroleum-based industry. Crystalline structure and association with lignin and hemicellulose makes lignocellulose recalcitrant towards enzymatic degradation. Owing to its recalcitrance, the efficient degradation of lignocellulose assumes well-balanced cooperation of different enzymes. The major components of lignocellulolytic enzyme cocktails are hydrolytic enzymes. However, recently discovered lytic polysaccharide monooxygenases (LPMOs) and their positive effect on lignocellulose degradation has led to intensive research of redox-active enzymes. For oxidative cleavage of cellulose, LPMOs need hydrogen peroxide cosubstrate. However, H2O2 has turned out to be a double-edged sword, being efficient co-substrate but also inactivating for LPMOs. Controlling the H2O2 levels is therefore of utmost importance in maintaining the stability of LPMOs during lignocellulose degradation. The present PhD project aims to find a suitable enzyme cascade and reaction conditions to control the levels of H2O2. Different H2O2- producing (oxidases), H2O2- consuming (LPMOs, peroxidases, peroxygenases), and H2O2 housekeeping (catalases) enzymes will be tested as the components of the enzyme cascade.

Supervisor: Angela Ivask

This project studies the stress responses and physiological adaptation of bacteria after their exposure to antimicrobial materials, and the role of the observed adaptations in the development of antibiotic resistance. It is known that unfavourable environment, including antimicrobials, induce various stress response pathways in bacterial cells. The outcomes may be physiological changes that lead to a higher resistance towards the specific unfavourable environment but may also increase bacterial tolerance or resistance towards other stressors. This project will utilise different types of antimicrobial materials with various modes of action and speeds, to follow their induced stress responses in bacteria. By linking the observed stress response pathways with bacterial antibiotic sensitivity proles we will draw conclusions on the potential of antimicrobial products to induce antibiotic tolerance and resistance in real-life situations.

Molecular Biotechnology

Supervisors: Tanel Tenson and Vasili Hauryliuk

Bacterial toxin-antitoxin systems (TAs) are diverse two-gene elements of genomes that are widespread in bacteria. The various protein toxins target different core processes of the cell in which they are encoded to dramatically inhibit growth, and the antitoxins efficiently neutralize the toxicity in a number of ways. In this project we will use bioinformatic, microbiological, biochemical and structural methods to understand the molecular basis of specific inhibition of toxins by antitoxins and uncover the mechanism of TA-mediated defence against bacteriohages.

Supervisors: Mart Loog, Oliver Lukason, Siim Salmar

Forest biorefining - the sustainable processing of (forest) biomass into a spectrum of marketable products and energy - has been the focus of significant research and development over the last years. The Wood Chemistry and Bioprocessing Laboratory of the University of Tartu, in collaboration with Barrus AS has designed a Ph.D. project focusing on the analysis of different forest bioeconomy models and various new technologies with the final aim to find the most optimal techno-economical solutions for forest biorefineries in the Baltic region. Barrus AS, a timber company located in southeastern Estonia, is one of the biggest pine window scantling producers in Europe. However, the production process leaves ca 75,000 tons of leftovers annually. Currently, most production residues go to the wood pellet industry and some fraction to the pulp manufactories. Because of that, Barrus AS is looking for innovative alternative solutions to valorize production leftovers to real high-valuable bioproducts. The project offers the Barrus AS a comprehensive assessment of technologies for deployment in the near term. The thesis will take a regional approach for the Baltic region to evaluate pathways for forest bioeconomy development. The project will perform a techno-economic analysis of most emerging technologies for production of biofuels, bio-based chemicals, or other biomaterials. The analysis assesses the regional performance of different conversion pathways and specific technologies. A case study will be applied to evaluate several options and a detailed feasibility study and market analysis will be performed for selected technologies to determine the best configuration. The project results will help Barrus AS achieve sustainability goals and be viable over the long term within the context of the regional forest, energy, and manufacturing sectors and global markets.

Supervisors: Ebe Merilo, Hanna Hõrak

Plant hydraulic status depends on soil water content and atmospheric vapour pressure deficit (VPD), representing supply and demand functions, respectively, for plant water status. During climate change, VPD increases, resulting in reductions in stomatal conductance, net assimilation rate and thus, production of plants. However, many aspects of stomatal VPD sensing and sensitivity are still vaguely understood. During this project, we will 1) study VPD-induced stomatal signalling pathway in Arabidopsis, using putative targets of OST1, a protein kinase crucial for stomatal VPD-induced closure. 2) clarify the mechanism of VPD-induced stomatal closure in barley using Arabidopsis-based knowledge about stomatal VPD regulation. In barley, laboratory VPD-experiments will serve to understand the mechanism of VPD-induced stomatal closure. Additionally, field experiments will be conducted together with Estonian Crop Research Institute, to study the correlations between gas exchange traits and grain yield of barley lines with different VPD-sensitivity. Understanding the molecular mechanisms underlying the stomatal VPD response and factors behind different VPD-sensitivity of plants is relevant to improve modelling outputs under different climate change scenarios and also, assist in breeding plants better adapted to future climate conditions.

Supervisors: Kaspar Valgepea, Kristina Reinmets

The quest to develop and adopt sustainable and carbon-free energy production technologies is key in tackling climate change. At the same time, the increasing emission of industrial waste gases and accumulation of solid waste possess a challenge for global biosustainability. Gas fermentation has emerged as an attractive technology for the conversion of waste feedstocks into value-added low carbon fuels and chemicals using acetogen bacteria. However, better understanding of carbon fixation and enzyme functionalities are needed to accelerate rational metabolic engineering of acetogens. This project aims to improve autotrophic carbon fixation in acetogens by in-depth functional characterisation and engineering of the carbon-fixating WoodLjungdahl pathway at both gene and protein levels. We predict our approach to advance systems-level understanding of gene functionalities in acetogens and accelerate their metabolic engineering into cell factories.


Supervisor: Manuel Hohmann

Recent observations of gravitational waves, the orbits of stars around the galactic center and the first image of M87*, have opened a new view on the cosmos. Assuming that gravity is described by general relativity (GR), it is concluded that these observations are explained by of black holes, which have been predicted theoretically as vacuum solutions to Einstein’s equations already in 1916. GR, however, has severe tensions with current cosmological observations, as well as with quantum theory, and so it is generally understood that it needs to be modified in order to encompass these tensions. A large class of such modifications is known as teleparallel gravity theories. These theories have received growing attention during the last decade, as they provide possible pathways to consistency of the theoretical description of gravity with both cosmological observations and quantum theory. While solving Einstein’s field equations in vacuum and spherical symmetry uniquely leads to the Schwarzschild solution describing a black hole, modified theories of gravity allow also for other solutions. Such solutions are denoted “exotic compact objects” (ECOs). Examples include wormholes, boson stars and gravastars. While some classes of ECOs can easily disguise as black holes in current observations, others have distinct signatures, for example in their gravitational wave pattern, which allows to distinguish them. The aim of this thesis is to study ECOs, their fundamental properties and possible observational signatures in modified teleparallel theories of gravity. In particular, we will study which types of ECO solutions with spherical symmetry exist in certain theories, derive these solutions, and discuss their properties, such as the required matter content in order to sustain them, as well as observational signatures in order to distinguish them from black holes using gravitational wave observations.

Supervisors: Vitali Nagirnõi, Dmitry Spasskiy

The project is aimed at the development of bright luminescence materials suitable for implementation in phosphor converted LEDs and detectors for remote luminescence thermometry. The main emphasis will be placed at the research of the correlation of structural and luminescence properties in molybdenum oxides K5RE(MoO4)4 and their solid solutions (RE = rare-earth element). It is expected that the structural disorder due to an incommensurately modulated crystal structure of some molybdates may attenuate the forbiddance of the RE 4f-4f transitions and prevent concentration quenching of RE luminescence thus facilitating a higher light yield. The level of the structural disorder can be tuned by varying powder synthesis or crystal growth conditions. Energy transfer to luminescence centre in RE based compounds as well as temperature dependent luminescence properties: thermal stability of emission, important for application in pcLEDs, and the redistribution of emission intensity between the Stark components of excited states of RE, essential for remote temperature detecting will be studied by time-resolved luminescence spectroscopy. The brightest compounds with suitable spectral and thermal luminescence properties will be selected for implementation studies.

Supervisors: Siiri Salupere, Rodrigo de Oliveira, Marti Jeltsov 

The PhD project will combine dose rate modelling of materials and technical evaluation of radioactive waste management options for small modular reactors (SMR). Nuclear power is an option for decarbonisation of the power sector in Estonia but it comes with certain field-specific challenges, e.g. waste management. This project studies the challenge from a technical perspective. The work provides valuable information to the national waste management authority responsible for the choosing the solutions for Estonia. On an international scale, an efficient model for radioactive waste management for nuclear newcomers opting for SMR technology is proposed. During the first years of the PhD project, a model will be developed using the EGSnrc software to assess attenuation of ionizing radiation in materials that can possibly be used in radioactive waste management. The model will be applied to innovative concrete mixtures containing oil shale ash and fibres. Addition of oil shale ash is expected to enhance the immobilization properties of concrete while adding fibres will enhance the mechanical characteristics. Modelling results will be validated by experimental determination of gamma ray attenuation with gamma spectrometric measurements. To evaluate the suitability of innovative concrete mixtures for SMR waste management, SMR waste source term and waste management pathways will be assessed. This doctoral project aims to propose the optimum waste management pathways per waste stream originating from an SMR. A less conservative model for dose assessment would allow for more adequate waste management solutions in terms of societal, technical, and economic efficiency.

Supervisors: Laur Järv, Aneta Wojnar

Various cosmological and theoretical issues motivate to consider gravitational theories beyond general relativity. Explorations in the sector where the affine connection is allowed to be independent of the metric in the geometric setup (where Palatini and teleparallel are perhaps the simplest types of models) have been receiving growing attention recently. At the same time the increasing amount of data regarding astrophysical objects as well as the new gravitational wave detections, allow to put the proposed models under comprehensive tests. First, the theoretical solutions obtained by solving the modified Einstein field equations need to be properly studied before applying them to the description of the physical objects. Healthy and realistic solutions will be then used to study compact objects such as black holes, neutron and white dwarf stars, with the particular focus on the cooling process of the dead stars. Moreover, such compact objects turn out to alter the propagation of the gravitational waves in their neighborhood. The change in the gravitational wave signal is different for a given theory of gravity, therefore new effects, which are not present in general relativity, can be studied.

Supervisor: Vijayakumar Anand

In the recent years, chaos-inspired imaging technologies (CI2 -Tech) have gained a lot of attention. Most of the CI2 -Tech approaches involve scattering of the light diffracted from an object to distinctly encode many spatio-spectral information channels in one intensity distribution. The doctoral project titled “Multispectral multidimensional imaging using an ensemble of selfinterfering spatially incoherent chaotic scalar waves,” unifies all the existing CI2 -Tech methods under a single roof. The doctoral project aims to create a computational optics framework for statistical optical experiments. In this framework, an ensemble of chaotic scalar waves with interesting intensity distributions can be synthesized, their composition and mutual interactions can be controlled. The project aims to access the non-linear regions of imaging characteristics by mapping every object point to an ensemble of special optical intensity distributions. When the constituents of the ensemble are sparse it is possible to achieve a strong dependency between the nature and composition of the ensemble and the imaging characteristics. When the density of the ingredients increases, the system collapses to the case of a regular scatterer. Consequently, the previous CI2 -Tech methods are only a special case in the proposed framework. The doctoral thesis aims to create a novel hybrid reconstruction algorithm based on cross-correlation, maximum likelihood and optimization to reconstruct the recorded intensity distributions into a 5D image. The doctoral project will generate new knowledge in statistical optics, reconstruction mechanisms and imaging technologies which has potential for developing novel microscopes and molecular fingerprinting sensors in mid-infrared wavelengths.

Supervisor: Marco Kirm

Scintillators are materials converting ionizing radiation and particles into low-energy radiation appropriate for photodetectors. The aim of our project is to create a path for new generation of ultrafast scintillators with time resolution in ps time domain. A concept of electronic band structure engineering will be applied to ternary compounds with complicated valence band structure, which create favourable conditions for appearance of intrinsic emissions, cross-luminescence and intra band luminescence, with short sub-nanosecond decay times. Suitable materials will be selected on the basis of open databases (e.g. AFLOW), consisting of results of theoretical studies of material properties, and experimental results available. Thereafter the selected materials will be synthesized as pure compounds and their solid solutions, their properties modelled and studied in home laboratory and at large scale facilities (MAX IV Lab, DESY Photon Science) using time resolved luminescence spectroscopy with sub-nanosecond time resolution at its top level. These materials will be beneficial for very different areas of technology, including high-energy physics (CERN), homeland security, controlled nuclear fusion research and especially medical diagnostics. Significantly improved time resolution will cause a breakthrough in positron emission tomography by allowing non-expensive low-dose scans for pediatric, prenatal and neonatal diagnostics in every major hospital, allowing early identification of tumour’s and other harmful conditions.

Supervisors: Hardi Veermäe, Joosep Pata

Gravitational-wave astronomy is a new and rapidly developing direction in experimental physics. The early Universe is transparent to gravitational waves. Thus it will soon become possible to observe gravitational-wave signals originating from processes of this epoch directly. To recognise these signals, it is crucial to connect the high energy theory with concrete experimental predictions. This is the primary goal of the doctoral project. It will focus on the numerical simulation of processes related to first-order phase transitions in the early Universe and the development of novel computational tools. The resulting predictions will improve our understanding of the first moments of our Universe.

Supervisor: Vijayakumar Anand

Interferenceless coded aperture correlation holography (I-COACH) is an incoherent holography technique capable of reproducing three-dimensional (3D) information of an object from a single camera shot. In I-COACH, the light from an object was modulated by a quasi-random phase mask and the scattered intensity pattern was recorded by an image sensor. The 3D object information was reconstructed by processing the object intensity pattern with the pre-recorded 3D PSF distributions. While I-COACH can record and reconstruct 3D information without two beam interference unlike its’ precursors such as self-interference digital holography methods, it is not without problems. The need for scattering in I-COACH and the need for recording PSFs increases the noise and reduces the resolving power respectively. The speckle distribution generated in I-COACH and the reconstruction mechanism involving cross-correlation with PSF precludes the introduction of special imaging characteristics. The doctoral thesis titled “Interferenceless coded aperture correlation holography with deterministic optical fields,” is a game changing approach which is expected to address the above challenges of I-COACH and expand the applicability of I-COACH to power sensitive areas. The proposed doctoral thesis will investigate the special beams with interesting spatial intensity distributions for 3D imaging applications. The doctoral thesis will create a unified artificial intelligence based image reconstruction algorithm for reconstruction of 3D information for special beams. The outcomes of the doctoral thesis is expected to lay the foundations of a new generation of imaging and microscopy technologies.

Science Education

Supervisors: Miia Rannikmäe, Regina Soobard, Jack Holbrook

This research is grounded on self-determination theory defining a person’s subjective beliefs about their successful task performance capabilities on testing students’ STEM competence including new components such as students’ understanding about science (NOS) and careers and on measuring students’ attitudes towards STEM learning, with stronger emphases on negative attitudes and on reasons why those appear.

The main goal is to determine the relationship between student STEM competence, perceived attitudes towards STEM education and self-efficacy related to STEM-career awareness, including future need competences for life issues, plus develop and validate an operational model for STEM education meeting society needs.

Research  data are collected among Estonian students and for the researcher good command in Estonian  language is obligatory.

Speciality admission requirements in Science Education :

  • the Doctoral project
  • admission interview

In the first stage of assessment, the doctoral thesis project will be assessed. In the second stage of assessment, there is an interview.

Both the doctoral thesis project and admission interview are assessed on a scale of 0 to 50 points, the minimum positive score is 35 points. Each components gives 50% of your overall score with a maximum overall score of 100 points. If you score at least 35 points out of 50 for the thesis proposal, you will be invited to attend an entrance interview.


Requirements for the content and form of the doctoral thesis. Here you will also find who to contact with questions.

The lenght of the doctoral thesis project is up to five A4 pages (without bibliography).

The project presents

·         the research challenge,

·         explains its innovativeness and importance, based on the scientific literature,

·         the research methodology proposed and the resources needed to address it,

·         describes how to expected results are relevant in practice and on the international scientific community.


The candidate is provided with information on the basis of which the candidate can get an overview of the candidate's research potential:

·         information on previous scientific activities (incl. working in positions containing research tasks),

·         a list of published scientific publications and presentations at conferences.




Doctoral thesis project assessment criteria (at least 35 points are required to fulfill the acceptance condition):

1. innovativeness and importance of the doctoral thesis (up to 10 p);

2. justification of the research problem (up to 10 p);

3. adequacy of research methodology (up to 10 p);

4. feasibility of the design (up to 5 p);

5. candidate's research potential for a design-based doctoral thesis (up to 15 credits).


During the admission interview, the candidate will interpret his/her doctoral thesis in English for 5 minutes, followed by an interview with the committee for up to 25 minutes.


Criteria for the evaluation the interview (at least 35 points are required to meet the admission requirement):

1. ability to dechipher the plan of the doctoral thesis (incl. research problem, related scientific literature, methodology) (up to 15 p);

2. motivation to study for a doctorate in education and to work in the field (up to 5 credits);

3. wider analytical and generalization skills in pedagogical topics (up to 10 p);

4. orientation in the problems of Estonian and world educational life (up to 10 p);

5. self-expression skills (all candidates have the ability to express themselves in English, candidates who speak Estonian as their mother tongue also have the ability to express themselves in Estonian) (up to 10 p).


Speciality representative: Miia Rannikmäe,

Space research and technology

Supervisors: Kersti Kangro and Krista Alikas

Phytoplankton is the basis of the food chain and important component of food-web and carbon cycle in all water bodies, being the first component in the lake to react to the changing environmental conditions. In the boreal region the changes in phytoplankton community are predicted due to more frequently appearing ice-less winters and higher summer temperatures favouring cyanobacterial growth. This means more potentially toxic cyanobacterial blooms, which are affecting the lake functioning and public usage. Remote sensing gives a possibility to get an overview about bloom parameters for the entire lake, multiple lakes simultaneously and with higher frequency than national monitoring possibilities offer. EU Copernicus programme satellites Sentinel 3/OLCI and Sentinel 2/MSI allow to monitor Chl a changes in boreal lakes, with an inclusion of Lake CCI database changes in temperature in various lakes can be studied. This doctoral thesis focuses to phytoplankton community changes detected from in situ and satellite data, using hyperspectral data and microscopically determined phytoplankton community composition to: -Study cyanobacterial bloom parameters from Sentinel 3/OLCI and Envisat/MERIS in shallow eutrophic lake Peipsi -estimate changes in phytoplankton community in larger boreal lakes -detect community changes from hyperspectral parameters gathered from various lakes seasonally during multiple years, based on Estonian and Swedish data.

Supervisors: Rain Kipper, Elmo Tempel

One of the main problems of astrophysics is the existence of dark matter, but not knowing what exactly it is. For solving it we need to study the dark matter as diversely as possible. A way with high potential, but not well explored is dynamical friction, which potentially allows to study the behaviour of dark matter on small scales. During this PhD project we study what conditions and where the effects of dynamical friction come forth best, select best candidates to study and study them from observations, and test what is the (relative) contribution from dynamical friction caused by baryonic matter and dark matter.

Supervisors: Antti Tamm, Lauri Juhan Liivamägi

Last decades have seen a huge progress in our understanding of the universe, largely thanks to the rapid development of observational facilities. According to the general undertanding we are living in a Big Bang universe, 69% of which constsits of dark energy, 26% of dark matter and a mere 5% of the normal baryonic matter, known to us. Unfortunately, numerous experiments and observational projects have not revealed, what dark matter is made of and what is the nature of dark energy. We have to admit that our theories of fundamental particles and gravity are either incorrect of incomplete. One key to solving this puzzle may lie in studying the largest structures in the universe, galaxy superclusters. In this project, data from the novel J-PAS cosmological survery will be used to map galaxy superclusters in the nearby and distant universe. The unique methodology applied in J-PAS grant an unprecedentedly deep and extensive astronomical dataset, with which we hope to clarify the role of large density pertubations in the formation and evolution of structures and check whether the current cosmological models support the existence and evolution of the largest supercluster complexes.

Supervisors: Mihkel Pajusalu, Rene Laufer

The goal of this doctoral thesis project is to investigate the selection of machine learning algorithms for space application, comparison of them with more traditional algorithms and investigation on how the specifics of space applications affect this. In addition, we will investigate the effect of the space environment on the use of such algorithms and how the choice of the algorithm can affect the architecture of the whole system. The main goal of this is improvement of space mission scientific output in limited bandwidth scenarios and overall reliability of such systems. The study will be conducted while developing the OPIC (Optical Periscopic Imager for Comets, launch planned in 2029, delivery to the prime contractor in 2025) instrument onboard the ESA Comet Interceptor mission as a real application scenario. For this, it has to be evaluated which kinds of algorithms and especially machine learning approaches would be applicable for this mission. The goal is to maximise the scientific return through the limited resources available. This will require the study of various tools, but also the development of own deployment tools to cover as many options as possible and to be able to evaluate all import design options. Ultimately, the chosen design will be tested in the OPIC scenario and potentially in applications in other fields, such as Lunar rover development, will also be explored.

Supervisors: Andris Slavinskis, Pekka Janhunen

Slavinskis and Janhunen have cooperated in electric sail demonstration since the ESTCube-1 project in 2011. They have been publishing ESTCube-1 mission design and results, ESTCube-2 mission design, FORESAIL-1 design, Aalto-1 design and results, and the MAT concept and design together. With the first two ESTCubes, the goal has been to demonstrate the E-sail in low Earth orbit (LEO) where it can also be used as a plasma brake for deorbiting. With ESTCube-3 and this project, we will close the gap between the LEO demonstration results and the future applications of the E-sail, such as MAT. This project will design the ESTCube-3 mission to demonstrate the electric sail in its authentic environment, the solar wind. Palos will improve and publish E-sail mission design tools and propose a secondary mission objective to flyby a near Earth asteroid.

Euro rahatähed_allikas Pexels

Research news: financial well-being, dialect training for computers, vitamin D, and parasitic helminths 

#research #for society
Pilt antibiootikumidest_allikas Pexels

Antibiotic resistance could be the biggest health crisis of this century