Thesis topics from Tartu Observatory

Superbubbles in galaxies are the results of many stellar explosions -- supernovae. The bubble pushes the gas away from the progenitor supernovae and forms a region devoid of gas. These are not well studied objects and more research in the topic is speeding up. We propose a project to see what stages most of the bubbles are and how many stellar explosions are needed to produce these objects.

Supervisors: Aikaterini Niovi Triantafyllaki and Rain Kipper

Lopsided galaxies are a possible new tool to study the dark matter distribution in the Universe, but only in case if the lopsidedness of the galaxy is caused by dynamical friction. Friction is expected when galaxy is moving and accelerating toward the heaviest object in the vicinity -- possibly a big galaxy cluster. With this thesis topic, we propose to study the opportunity if lopsided galaxy tail could be directed to the closest clusters and if the dynamical friction could be the culprit of lopsidedness.

Supervisors: Rain Kipper and Aikaterini Niovi Triantafyllaki

The environment around the galaxy is pushing the gas away creating tails of stars, gas and dust, giving them the image of a "jellyfish". The shape of the tail specifies what the galaxy went through. In this project we want to study the positions of regions for star formation in these tails.

Supervisors: Aikaterini Niovi Triantafyllaki and Rain Kipper

Galaxy bars and lopsidedness are both signs of the galactic disc being unstable. Mostly, these non-stabilities are treated as independent, but there is no evidence this is the correct approach for this studies. In this thesis topic we propose to investigate through image processing and statistics the galaxy bars and lopsided levels and to see if these correlate.

Supervisors: Rain Kipper and Aikaterini Niovi Triantafyllaki

Density fields are a way to describe distribution of matter in the Universe. They can be useful tools for characterising the overall large-scale structure or study the dependance of galaxy properties on the environment. There are several methods to calculate density fields using both observational data and N-body simulations. The aim of the thesis would be to test and compare different algoriths.

Supervisor: Juhan Liivamägi

Despite making up 85% of the matter content, little is known about what DM is. Your thesis will be concerned with exploring the power of observations of our own Galaxy to constrain the nature of DM. The exact project will depend on your interests. It may be more theoretical work on primordial black holes and their detectability with the next generation of gravitational wave detectors (such as ESA/LISA) or dedicated to developing machine learning approaches to detect dynamical effects of DM subhalos. For more information, please contact me directly.

Supervisor: María Benito

4MOST is the next-generation astronomical observing project aimed at observing tens of millions of stars, galaxies, supernovae, and other interesting astronomical objects over a five-year observation period. The Tartu Observatory team is actively involved in preparing the 4MOST observing program and contributes to ensuring that the 4MOST observational data are suitable for various statistical analyses. To understand and analyse observational selection, we use the 4MOST observation simulator developed at the Tartu Observatory. In connection with the 4MOST project and data from observations, the Tartu Observatory offers a variety of bachelor's and master's thesis topics. The work involves conducting simple statistical analyses based on observational data and visualising the observational data. Some programming skills and experience in data visualisation (such as in the Jupyter Notebook environment) would be advantageous. The specific content and degree of difficulty of the work are determined in cooperation with the student. For more information, please contact the supervisors.

Supervisors: Elmo Tempel and Taavi Tuvikene

The distribution of galaxies in the universe can be described as a cosmic web, with galaxy groups and clusters at the nodes connected by galaxy filaments. Future observing programs will allow us to see the cosmic web in ten times more detail than currently available data. 4MOST is one such observing program that reveals the fine details of the cosmic web. Existing tools must be further developed and enhanced to investigate the highly detailed cosmic web. The Tartu Observatory has extensive experience detecting galaxy clusters and filaments from observational data. To stay at the forefront of science, existing methods must be continually improved and developed further. Students can also contribute to this process. If you are interested in the cosmic web, please get in touch, and we can discuss various ongoing projects in more detail. The work requires an interest in programming and the development of various algorithms.

Supervisors: Elmo Tempel and Taavi Tuvikene

One of the biggest challenges in observational cosmology is the determination of galaxy distances. The accuracy of distance measurements (redshifts) using spectroscopic observations is quite good, but unfortunately, the number of galaxies that have been spectroscopically observed is limited. On the other hand, the number of photometrically observed galaxies is several orders of magnitude larger. The distances to these galaxies are then measured utilising photometric redshift methods. One of the realisations of this method has been developed in Tartu Observatory - Tartu Observatory Photometric redshifts (TOPz). In addition to the measured distances, TOPz enables us to evaluate other galaxy properties (stellar mass, star formation history, etc.). In this project, you would apply the TOPz code to the observed photometric galaxy data and determine their physical properties in order to understand galaxies' evolution as a population. This work involves some simple statistical analysis and the making of various plots. Some prior knowledge of Python and jupyter notebooks would be beneficial. If you want more details about the project, then contact us, and we would be happy to talk more.

Supervisors: Elmo Tempel and Jaan Laur

The evolution of galaxies is an intricate process shaped by various factors, including the influx of gas, the presence of spirals, and the existence of supermassive black holes at the galaxy's core, along with galaxy interactions. Simultaneously considering these complex aspects poses a challenge, influencing the evolution of distinct regions within a galaxy, e.g. the bulge, disk, and halo. To simplify our understanding of galaxy evolution, we adopt a step-by-step approach, dissecting galaxies into components (such as disks and bulges) and scrutinising their individual development. We propose an engaging thesis project that involves image processing to identify and analyse different galaxy components. This work forms a crucial part of a larger project aimed at unravelling the cosmic intricacies of galaxy evolution.

Supervisors: Elmo Tempel and Rain Kipper

Fossil galaxy groups are galaxy groups with a mostly very bright elliptical galaxy at their centre. The other members of the group are much less luminous than the central galaxy. In most cases, the fossil group is surrounded by a large X-ray halo. Fossils are assumed to be poor groups in which most of the members have merged with the central galaxy. The aim of the study is to compare the positioning of the fossil groups and the reference sample in the cosmic web in order to better understand the evolution of the groups.

Supervisor: Maret Einasto

Emission-line imaging with radio telescopes is routinely used to assess the cold gas content (atomic and molecular) of galaxies. This helps in studying how changes in gas reservoirs drive galaxy evolution. Such imaging reveals how the environment in which galaxies are embedded affects their gas content through processes such as tidal interactions, active stripping, or gas accretion. For a more complete understanding of galaxy evolution, it is important to extend these measurements to more distant galaxies. However, this is challenging due to the inherent weakness of emission lines and the presence of comparatively strong man-made interference.

The proposed thesis will contribute to a larger project aimed at developing neural network–based signal detection software. The student will have the opportunity to choose which parts of the project to engage with. This will enable them to gain experience in working with radio observations, using software for data processing, and applying advanced statistical analysis, signal processing, and machine learning techniques. The project will require programming in Python.

Supervisor: Boris Zhivkov Deshev ([email protected])

Image

Leaf angle distribution (LAD) is a key parameter in models useful for understanding vegetation canopy processes of photosynthesis, evapotranspiration, radiation transmission, and spectral reflectance. Yet, despite the strong sensitivity of many models to variability in LAD, the difficulty in measuring LAD causes it to be one of the most poorly constrained parameters. This doctoral thesis will devise, test, and implement a novel methodology exploiting multi-angle sensors to retrieve LAD information.

Read more.

Supervisor: Jan Pisek

Image

Assessing biodiversity through ecological field data encounters various challenges, particularly in gathering reliable information for large areas. There is a pressing need for operational techniques utilizing remotely sensed data to aid ecologists and decision-makers. This doctoral thesis would explore spectral diversity metrics, derived from optical imagery, and machine learning for spatially predictive biodiversity modelling.

Read more.

Supervisors: Jan Pisek and Evelyn Uuemaa

Image

Needle shoot architecture for given tree species may change due to various factors, including age, position, orientation, and geographical location. Using high-precision blue light 3D photogrammetry scanning, the aim of this MSc. thesis would be to scan needle shoots from different positions and locations and explore how the architecture may or may not change.

Supervisors: Jan Pisek

Supervisor: Kersti Kangro

Supervisor: Lea Hallik

Supervisor: Mait Lang

Supervisor: Krista Alikas

Supervisor: Krista Alikas

Supervisor: Krista Alikas

Supervisor: Erko Jakobson

Modern terrestrial laser scanners provide point clouds where spacing is 2-3 mm at 10 m distance from scanner. Such point clouds contain useful information about growing trees in forests. Construction of mathematical models of trees by fitting geometric shapes on the point clouds is one solution for extracting the information. Although general solution pathways and even preliminary software exists, much problems are unsolved regarding precision and speed of the methods. What would be optimal solution for hemiboreal forests in Estonia?

Supervisor: Mait Lang

Public access is granted to use data collected with environmental monitoring satellites Landsat-8&9 OLI-TIRS and Sentinel-2 MSI. The multispectral images have good radiometric resolution and are cost efficient data sources for forest inventories. Interpretation of the extracted spectral signatures can be done using forest radiative transfer models (FRT). Which variables of forest structure have strongest influence on the forest spectral signature?

Supervisor: Mait Lang

Multispectral Landsat-8&9 OLI-TIRS ja Sentinel-2 MSI time series record plant canopy properties via recorded spectral reflectance. Which are the main types of time series and how large is the variability of vegetation parameters within classes? The knowledge will help to advance monitoring of plant vegetation status and select informative descriptive variables for machine learning algorithms.

Supervisor: Mait Lang

Assessment of forest inventory variables using active space-borne microwave sensors.
Content: Radars emit microwave pulses and register the backscatter. Microwave remote sensing using space-borne radars are substantially less dependent on clouds compared to sensors that measure radiation in the optical part of spectrum. Bistatic radar satellite constellations like TanDEM-X provide measurements that contain spatial information about forest canopies. However, it is clear if there exists a universal model for extracting the estimates of forest canopy height, canopy cover and biomass and other forest inventory variables. Are the predictions dependent on growth season and tree species composition?

Supervisors: Kaupo Voormansik, Mait Lang

Enabling artificial intelligence processing for data reduction and autonomous navigation onboard spacecraft for exploration missions. Improvement of data processing speed and efficiency through dedicated hardware and programmable logic.

Supervisor: Ric Dengel

Image

Development of autonomous traversals functionality for the Kuupkulgur lunar rover project aiming to bridge the gap between local planning and execution of rover manoeuvrability in preparation for future lunar rover missions. The development work is in support of the Kuupkulgur project and the expected results are to be verified and validated on the rover platform utilizing the Lunar analogue site at Tartu Observatory.

Supervisor: Quazi Saimoon Islam

Image

Development of a light source that simulates closely the properties of solar illumination in a controlled environment. The proposed light source is required to closely replicate the effects of solar illumination in the testing facilities at Tartu Observatory, with a complete evaluation of its physical accuracy and properties.

Supervisor: Quazi Saimoon Islam

Environment mapping is a crucial step for autonomous rovers. Most of the research in this field targets either indoor or urban environments. Our workgroup's research aims to bring it closer to off-road use cases, for which we are investigating the use of deep learning. This research focuses on the design, development, and evaluation of a critical step in the pipeline: training data generation. There are many options, and the specific topic will be tailored to fit your best interest.

Supervisor: Rando Avarmaa

Supervisor: Mihkel Pajusalu

Supervisor: Laila Kaasik

Image

Dynamic events involve rapidly changing states, information about which is essential to understand their origin and impact. This research is to develop strategies to be able to map and characterize such events. The research can be applied to mission designs to survey and represent events occurring on terrestrial and extra-terrestrial surfaces.

Supervisor: Aditya Savio Paul

High-mass stars typically evolve into red supergiants during their lifetime and stay in this stage of evolution for quite a long time in their lifetime. In general, all stars evolved away from the main sequence are variable to a greater or lesser extent, the timescale and amplitude of variability changing during evolution. The aim of the work is to find out whether there are stars with clearly distinct variability characteristics among selected red supergiant stars that are similar in terms of absolute luminosities and photometric colors, which would indicate their different stages of evolution as red supergiants. The research can use archival observations from Tõravere and the wider world, both from ground- and space observatories.

Supervisors: Indrek Kolka, Tõnis Eenmäe

Stars are often classified by their brightness and color, and many stars do not have a proper spectral classification. The distances of many high luminocity stars are not well known, even from Gaia measurements. Thus, for some types of stars located at the upper end of the Hertzsprung-Russell diagram, confusion is possible, where rather lower mass post-AGB stars are confused with more massive red supergiants or with stars in very exotic stages of evolution (e.g. hypergiants). The aim of the work is to find out the nature of candidate stars after the red supergiant phase found on the basis of Gaia DR3 data. In the course of the work, it is possible to use archive data as well as make observations using telescopes at Tõravere or abroad. See: https://doi.otg/10.1093/mnrasl/slac088.

Supervisors: Indrek Kolka, Tõnis Eenmäe

The Ariel space telescope will observe the atmospheres of exoplanets while they transit their parent star, as seen from Earth. In order to use the time of the space telescope as effectively as possible, it is necessary to know the exact ephemeris of the transits of the planets - the moments of the central moments of the eclipses. The depth of eclipses in the light curves is up to a few percent, usually considerably less. Although very good ephemerides have already been determined for many easily observable exoplanets, large telescopes are needed to accurately determine the ephemerides of lower apparent brightness parent stars or shallow eclipses. The aim of the work is to measure the transit parameters of some exoplanets from the Ariel object list with poorly known ephemeris using the new photometer of the Tõravere 1.5-meter telescope.

Supervisor: Tõnis Eenmäe

The Ariel space telescope will observe the atmospheres of exoplanets when exoplanet is transiting its parent star when seen from Earth. In order to make the most efficient use of the space telescope's time, it is necessary to know the exact ephemeris of the planetary transits - the time of the central moments of the stellar eclipses caused by their planets. The eclipse depth of transits in the light curves of the star is up to a few percent, usually much less. Many easily observable exoplanets already have very good ephemerides, but larger telescopes are needed to determine the ephemerides of low apparent luminosity parent stars or transits of shallow depth . The aim of this work is to measure the transit parameters of some of the exoplanets on the Ariel list with poorly known ephemerides using the new photometer on the 1.5 m telescope at Tõravere.

Supervisor: Tõnis Eenmäe

Before starting to use any new piece of scientific equipment, it is crucial to know the actual parameters of the device and the quality of the data it is able to provide. This is why most of the scientific instruments undergo a commissioning period, during which it is verified whether the instrument is working correctly and giving correct results. Very recently, two new instruments were built for the 1.5-m telescope in Tõravere and the telescope control system was significantly upgraded, too. All these instruments will require a significant number of tests and analyses before they can be put into routine operation.

With the launch of the instruments, it is possible to flexibly choose from a wide range of thesis topics in physics, metrology, computer hardware, e.g. (but not exhaustively) in the following broader areas:
New spectrograph (basic properties (including limiting magnitude), quality, stability, ....).
New photometer (calibrations, stability, ....).
The telescope itself (e.g. refinement of pointing, ....).

Supervisor: Tõnis Eenmäe

Many types of stars eject matter during the late stages of their lives. These outbursts can either lead to almost a complete destruction of the star, as in the case of supernovae, or, in lower-mass stars, occur as recurring events that help shed excess material. Examples of the latter include nova explosions and outbursts in symbiotic binaries.

The ejected material often forms intricate and beautiful nebulosities with complex, non-spherical morphologies which are mostly glowing due to the ultraviolet radiation of their central stars. While stars are mostly spherical during their stable phases, the remnants left behind after these outbursts often exhibit a variety of shapes. Understanding the mechanisms that shape these remnants is an ongoing challenge in astronomy, and no definitive explanation has yet been reached.

Researchers study these shapes using both astronomical images and spectral analysis. Students can explore this fascinating area by choosing a stellar remnant of interest from predefined list of objects for which the data is already available. Notable examples include the nova remnant GK Persei and the nebulosities surrounding the symbiotic binary R Aquarii.

Supervisor: Tiina Liimets