University of Turku Astronomy and Space Physics Master’s Thesis Guide: 30 ECTS, Observations, Modelling and UTUGradu

What this guide covers

This guide is for the University of Turku Astronomy and Space Physics track in the current 2024-2027 Exact Sciences curriculum. It uses the exact Peppi object EXACTASMDP2427, programme 97189, and keeps the thesis module separate from the thesis itself. The programme has a 40 ECTS Master’s Thesis Module, but the thesis component is 30 ECTS. Peppi lists TÄHT7042 Master’s Thesis, Astronomy and FFYS7070 Master’s Thesis, Physics as 30 ECTS thesis course routes, alongside MTDK1307 maturity, TÄHT7043 seminar and separate project or internship studies. The guide also explains how observational data, statistical inference, simulations, spectroscopy, high-energy astronomy, radio interferometry, heliophysics and space instrumentation should be bounded in a thesis.

Current programme object

The current Peppi programme object is EXACTASMDP2427 / programme 97189 for 2024-2027. The degree is Master of Science, organised in the Faculty of Science and Department of Physics and Astronomy. The formal degree size is 120 ECTS. Astronomy and Space Physics is one of six Exact Sciences specialisation tracks. The public programme describes three broad study lines: theoretical astrophysics, observational astronomy and space physics. Use the curriculum period attached to your own study right and HOPS rather than relying on an older Physical and Chemical Sciences structure, because the programme name and module organisation have changed over time.

How the 120 ECTS degree is structured

The current programme uses five broad parts. Advanced Studies are 80 ECTS and contain a 40 ECTS Master’s Thesis Module, a 20 ECTS Theoretical Courses in Space Sciences module and a 20 ECTS Methodological Courses in Space Sciences module. A 20 ECTS Thematic Module can be selected from options such as Special Courses in Space Sciences, Instrumentation for Space Sciences, Physical Modelling or Machine Learning. The remaining 20 ECTS are Other Studies, including obligatory Finnish-language studies. The public programme page likewise describes the compulsory thesis component itself as 30 ECTS.

Read the 40 ECTS Master’s Thesis Module correctly

The 40 ECTS module is not a 40 ECTS thesis. Peppi lists two 30 ECTS thesis course units in the module: TÄHT7042 Master’s Thesis, Astronomy and FFYS7070 Master’s Thesis, Physics. The programme-level requirement is a 30 ECTS thesis component, so these should not be added together as a 60 ECTS thesis requirement. The same module also includes MTDK1307 Degree Qualifying Examination at 0 ECTS, TÄHT7043 Seminar in Astronomy and Space Physics at 2 ECTS, FFYS7039 Topical Project in Research at 2-10 ECTS and FFYS7040 Internship at 0-6 ECTS. Verify which thesis course route applies to your registration.

TÄHT7042 Master’s Thesis, Astronomy

TÄHT7042 is a 30 ECTS astronomy thesis route. Its official learning outcomes emphasise becoming involved in astrophysics research, using scientific literature, developing observational and/or theoretical research skills, following good and ethical scientific practice, and presenting results clearly in relation to relevant scientific questions. The course is based on the student’s own research and relevant scientific literature. This makes a useful standard for the manuscript: the thesis should not be only a literature summary or only a technical workflow. It should show how a scientific question was investigated and how the resulting evidence was interpreted.

FFYS7070 Master’s Thesis, Physics

FFYS7070 is also a 30 ECTS thesis course in the same programme module. It trains independent laboratory or theoretical work, analysis and critical evaluation of results, scientific-literature use and good scientific practice. The course description says the thesis usually includes original experimental or theoretical work but can also be literature based. The thesis route therefore depends on the registered subject and project context. Do not assume that an Astronomy and Space Physics student must complete both TÄHT7042 and FFYS7070. The programme-level thesis component remains 30 ECTS.

MTDK1307 maturity examination

MTDK1307 is the 0 ECTS Degree Qualifying Examination. In the normal route, the thesis abstract or another suitable part of the thesis can function as the maturity test and demonstrate knowledge of the thesis field. A conditional written route applies only in particular circumstances involving prior degree and Finnish or Swedish educational-language background. Because the route is student-specific, confirm your own requirement in the current HOPS and University instructions. Do not copy another student’s maturity process without checking whether the same educational-language conditions apply to you.

TÄHT7043 seminar and other module studies

TÄHT7043 Seminar in Astronomy and Space Physics is a separate 2 ECTS course that trains scientific presentation through seminar participation and an own presentation. FFYS7039 Topical Project in Research is a separate 2-10 ECTS project that can be experimental, theoretical or literature based and requires a report. FFYS7040 Internship is another separate module object at 0-6 ECTS. The programme description characterises the 40 ECTS thesis module as a 30 ECTS thesis plus other studies such as internship or project work. These credits support the research environment but do not increase the thesis itself above 30 ECTS.

Choose the thesis mode before choosing the software

Astronomy and Space Physics supports very different thesis modes. An observational thesis may acquire or use telescope or archival data. A theoretical thesis may derive or numerically solve physical models. A space-physics thesis may analyse in-situ or remote-sensing measurements. An instrumentation thesis may design, calibrate, simulate or test a detector or instrument. Define the research question and evidence level before selecting software. Python, pipelines, Monte Carlo tools, telescope archives or instrument simulators are means to answer a question, not the research contribution by themselves.

Build a thesis plan around the evidence chain

The plan should define the research question, data or model source, physical assumptions, method, calibration or validation strategy, statistical treatment, uncertainty, expected outputs, risks and milestones. Observational work may depend on telescope allocation, archive availability, weather, calibration files or pipeline compatibility. Simulations may depend on compute resources, convergence and parameter exploration. Instrument work may depend on hardware, laboratory access or calibration sources. A written plan helps expose these dependencies early and gives supervisors a concrete basis for checking whether the intended conclusion is supported by the proposed evidence.

Use supervision as evidence control

Regular supervision is especially important when a project mixes astrophysics, statistics, coding and instrument or simulation work. Discuss changes in sample selection, calibration, model assumptions, priors, data-quality cuts, software versions, numerical parameters and excluded observations. If the method changes after preliminary results, record why. A supervisor can help distinguish a scientifically necessary change from a result-driven adjustment that could bias inference. The final Methods section should make the main decision path reconstructable rather than presenting the final pipeline as if it had been fixed from the beginning.

Observational data have several evidence layers

An observational thesis should distinguish raw observations, calibration data, reduced products, derived measurements and physical interpretation. Telescope data do not become scientifically neutral simply because a standard pipeline was used. Instrument response, flat-fielding, wavelength calibration, background subtraction, seeing, sky conditions, detector effects and data-quality flags can all influence the result. If archival data are used, preserve archive identifiers and selection criteria. A final catalogue value or spectrum is easier to trust when the thesis shows how it was produced and which systematic limitations remain.

TÄHT7055 and professional telescope observations

TÄHT7055 Observational Techniques Using the Nordic Optical Telescope trains students to plan and conduct professional observations, reduce and calibrate optical and near-infrared imaging or spectroscopy, analyse the data and report results in relation to an astrophysical question. A thesis using similar observations should document observing mode, instrument, filter or spectral setup, exposure strategy, calibration files and data-reduction choices where they affect the conclusion. If weather or technical problems cause archival data to replace planned observations, state this clearly because it changes the provenance and independence of the dataset.

ESO-style data processing and pipeline limits

TÄHT7056 covers modern imaging and spectroscopy, data reduction, instrumentation and extracting meaningful measurements from complex ESO-type datasets. Pipelines are valuable because they standardise processing, but a pipeline output is not automatically a validated physical measurement. The thesis should identify important pipeline settings, calibration versions and quality controls. If custom processing is added, explain why and test whether it changes results. When multiple reductions are possible, report enough information that an examiner can understand whether the scientific conclusion depends on a particular processing choice.

Calibration is not the same as absence of systematic error

Calibration corrects known instrumental or observational effects using reference information, but it cannot guarantee that every systematic error has disappeared. State the calibration model and its limits. If a zero point, response curve, background model, detector gain or point-spread function is estimated from data, include its uncertainty when it materially affects the result. A result that is statistically precise can still be systematically biased. Keep random uncertainty and systematic uncertainty conceptually separate rather than combining them into one vague error statement.

Statistical inference should match the scientific question

TÄHT7023 Statistical Methods covers uncertainty, confidence levels, least squares, goodness of fit, decision making and Bayesian analysis. Choose statistical tools because they answer the research question and match the data-generating process. Define the null or comparison model when significance is reported. If Bayesian inference is used, document priors and the likelihood. A statistically significant result is not automatically astrophysically important, and a best-fitting model is not automatically the true physical model. Interpret parameter estimates and intervals within the assumptions of the model and data.

Goodness of fit is different from model truth

A model can fit one dataset well while still being incomplete or non-unique. Evaluate residuals, uncertainty and goodness of fit rather than reporting only the best-fit parameter values. If several models describe the data similarly, say so. When a physical parameter is strongly model dependent, separate the directly measured observable from the inferred quantity. This is especially important for compact-object masses or radii, source temperatures, abundances, spectral components and transport parameters, where the mapping from data to physical parameter depends on an explicit model.

Fourier and time-series analysis need sampling controls

FFYS7010 covers Fourier transforms, discrete transforms, correlations and time-series applications in astronomy. A periodogram peak can be affected by sampling cadence, window functions, aliasing, gaps and preprocessing. Before claiming periodic or quasi-periodic behaviour, examine whether the feature could arise from the observing pattern or analysis choices. Filtering can change amplitudes and timescales, so document it. If multiple frequencies were searched, account for the search procedure when interpreting significance. A clean Fourier peak is not by itself proof of a physical clock in the source.

Signal and image processing can alter the evidence

FFYS7110 includes filtering, smoothing, convolution, deconvolution, image reconstruction, PSF matching, image subtraction, segmentation, registration and data fusion. These operations can increase scientific usefulness but also create dependencies on algorithm choices. Report thresholds, kernels, PSF models, registration references and reconstruction assumptions when they affect results. A visually sharper image is not automatically more accurate. If object counts, morphology or fluxes depend on processing parameters, use sensitivity checks or alternative settings to show whether the scientific conclusion is stable.

Simulation work needs both numerical and physical validation

FFYS7068 Classical Simulation Methods covers numerical computing, optimisation, Monte Carlo methods, particle models, molecular dynamics, continuum models and FEM. A simulation thesis should state equations, initial and boundary conditions, parameter values, numerical scheme, grid or particle resolution and convergence checks relevant to the claim. Numerical convergence only shows that the numerical result stabilises under the chosen formulation; it does not prove the physical formulation is valid. Compare with analytical results, benchmark problems or independent observations where possible.

Monte Carlo uncertainty should be labelled correctly

Monte Carlo methods can quantify sampling distributions or propagate uncertainty, but different uncertainty sources should not be mixed without explanation. Distinguish finite-sample variation, measurement error, parameter uncertainty and physical-model uncertainty. Report the number of samples or realisations where it matters and check whether results are stable. If a Monte Carlo model is calibrated using the same observations later used to judge performance, do not call that an independent validation. A large number of simulations does not compensate for a poorly specified physical model.

Hydrodynamics and plasma physics use different approximations

FFYS7090 Hydrodynamics covers ideal and viscous flows, gas dynamics, waves, instabilities and turbulence. FFYS7092 Plasma Physics ranges from single-particle motion and kinetic theory to fluid and magnetohydrodynamic descriptions. These approximations have different domains of validity. A thesis should identify the regime assumed and explain why it is appropriate. Do not move casually between kinetic, fluid and MHD conclusions as though they describe identical scales and processes. When the conclusion depends on an approximation, state this as part of the evidence boundary.

Radiative-transfer inference is model dependent

TÄHT7010 Radiative Processes in Astrophysics covers emission, absorption, scattering, radiative transfer and high-energy radiation mechanisms. A thesis using radiative modelling should identify geometry, source function, opacity, scattering assumptions and relevant particle distributions. Matching a spectrum does not prove a unique physical scenario because different parameter combinations can sometimes generate similar observables. If degeneracy exists, report it. Keep directly observed fluxes or spectra separate from derived temperatures, densities, magnetic fields or particle distributions that depend on the adopted radiation model.

Stellar models have their own validity limits

TÄHT7061 Stellar Structure and Evolution connects observations to hydrostatic equilibrium, energy generation and transport, stellar timescales, degeneracy and compact remnants. When stellar models are used in a thesis, state composition, evolutionary assumptions and model grid. Agreement with one HR-diagram location does not establish a unique evolutionary history. If age, mass or evolutionary state is inferred from a model grid, report the model dependence and observational uncertainty. The conclusion should not appear more precise than either the measurement or the stellar model permits.

Spectroscopic diagnostics require line-formation assumptions

TÄHT7053 Spectroscopic Diagnostics in Astrophysics uses line emission and absorption to estimate quantities such as temperature, density, composition and velocity. These inferences depend on atomic data, line formation, excitation and ionisation, extinction and whether LTE or non-LTE assumptions are used. A line identification is not the same as a quantitative abundance measurement. If blending, extinction correction or continuum placement affects the result, report it. Derived physical conditions should be presented as model-dependent inference rather than direct readings from the detector.

High-energy astrophysics separates observables from compact-object parameters

TÄHT7036 covers compact stars, accretion, pulsars, black holes, relativistic jets and high-energy sources. Many thesis questions in this area infer mass, radius, spin or accretion properties from spectra, timing or light curves. Keep the observable and the model-dependent parameter constraint separate. If different models produce different physical parameters, this is part of the scientific result rather than a nuisance to hide. State energy bands, background treatment, instrument response and timing selection when they affect the inference.

X-ray and gamma-ray pipelines need instrument context

TÄHT7058 teaches X-ray and gamma-ray instrumentation, archive search, ESA/NASA pipelines, spectral and timing analysis and observing proposals. A thesis using these data should preserve mission, detector, observation identifier, processing version and selection filters. Pipeline-generated spectra or light curves remain instrument-dependent measurements. Background regions, response files, pile-up controls, dead time or event filtering can matter. If multiple missions are combined, differences in calibration and bandpass should be addressed before interpreting apparent source changes.

Radio interferometry requires reconstruction awareness

TÄHT7062 Radio Astronomy and Interferometry covers single-dish work and interferometric imaging. Interferometric images depend on uv coverage, calibration and reconstruction. Angular resolution and sensitivity can vary across observations. A structure seen after imaging is not automatically a direct map of the sky without considering sidelobes, weighting and deconvolution. When comparing morphology or flux density across epochs or arrays, use compatible imaging choices or quantify the effect of differences. Report non-detections through sensitivity limits rather than treating them as proof of physical absence.

Heliophysics needs careful event association

FFYS7109 Heliophysics covers solar structure, solar wind, heliosphere, flares, coronal mass ejections and solar energetic particles. Space-physics theses often combine local particle measurements with remote observations and physical models. These are different evidence sources. Temporal association does not automatically prove a transport or acceleration mechanism. Define event-selection criteria, propagation assumptions, spacecraft location and data gaps. When comparing events, account for detection thresholds and selection effects so the conclusion is not driven only by the easiest events to observe.

Space Missions and Space Environment sets design constraints

FFYS7098 covers orbital mechanics, spacecraft subsystems and environmental conditions such as vacuum, temperature, radiation, plasma, debris and meteoroids. In a mission or instrumentation thesis, scientific requirements should be translated into mission and subsystem requirements. A feasible orbit or subsystem concept does not by itself demonstrate scientific performance. Environmental assumptions should be explicit because they can control shielding, thermal design, pointing, power, communication and detector behaviour. Separate mission feasibility, engineering design and scientific measurement performance as distinct evidence levels.

Space Instrumentation: trace science requirements into hardware

FFYS7105 covers spacecraft instruments, mechanical/electrical/data interfaces, timing, magnetometers, particle detectors, plasma probes and mission-driven instrument design. An instrumentation thesis should trace the scientific quantity to the detector or sensor, electronics, timing, calibration and data product. A prototype can demonstrate design feasibility under tested conditions but is not automatically a flight-qualified instrument. If the work uses simulations of detector response, distinguish simulated response, laboratory calibration and in-flight performance. Each supports a different level of claim.

Optical-system design is not as-built performance

FFYS7101 Optical Systems Applications covers aberrations, optical design, cameras, spectrometers, adaptive optics and integrated instrument design. Optical simulation can predict nominal performance, but as-built behaviour also depends on manufacturing tolerances, alignment, detector response and environmental conditions. A thesis should state whether a result comes from design software, laboratory measurement or on-sky performance. If tolerance analysis is important, include it. Do not describe a nominal optical design metric as measured instrument performance unless the instrument was actually characterised.

Instrument characterisation needs traceable references

FFYS7106 can involve designing, building or characterising space-science instruments as well as modelling or literature work. When a detector or instrument is characterised, define the calibration source, reference standard, geometry, environmental condition, repeatability and uncertainty where they affect the result. If a response curve is derived, show how raw counts became a calibrated response. If simulation and measurement are compared, keep them labelled separately. Agreement over one range does not prove accuracy outside the tested energy, wavelength or environmental range.

A prototype is not space qualification

A functioning laboratory prototype is important evidence for integration and basic performance, but space qualification normally involves additional environmental, mechanical, thermal, radiation, reliability and interface requirements. Unless such tests are actually part of the thesis, state that the prototype was demonstrated or characterised under laboratory conditions. This avoids turning a legitimate engineering result into an unsupported claim about flight readiness. The same principle applies to optical prototypes, particle detectors, calibration units and software-controlled autonomous instruments.

HPC improves capacity, not scientific validity

FFYS7111 teaches use of modern supercomputers, Linux shell workflows and batch-job systems. HPC can make parameter studies, large simulations and data processing feasible, but more compute does not make a physical model more correct. Preserve code, input data, environment, job configuration and random seeds where they affect results. If parallelisation or GPU execution changes numerical precision or algorithm behaviour, quantify the effect. Reproducibility depends on a traceable computational environment, not merely on recording that a supercomputer was used.

Keep observational and model provenance reproducible

Preserve observation IDs, raw or archival data sources, calibration files, reduction scripts, software versions, model code, parameter files, notebooks and figure provenance. A final PDF plot is not enough if the route from source data to scientific result cannot be reconstructed. For computational work, record numerical settings and software environments. For observational work, keep data-quality selections and calibration decisions. Reproducibility does not mean every dataset must be publicly redistributed, but an examiner should be able to understand how each major result was produced.

Company, collaboration and research-group boundaries

Many projects are embedded in active research groups or instrumentation collaborations. Clarify which data, code, hardware or preliminary results can be included in the public thesis. Collaboration does not reduce the need to identify the student’s own contribution. If a team supplies a calibrated data product or simulation framework, describe what was inherited and what the thesis independently changed, tested or inferred. Resolve publication, embargo or confidentiality issues before final submission rather than after sensitive material has been inserted into the manuscript.

AI use must remain accountable

University guidance on responsible AI does not transfer responsibility away from the student. If AI assists with code, literature triage, translation, plotting or drafting, verify the result and follow current University and supervisor instructions. Do not allow a generative system to invent references, observation identifiers, calibration values, physical constants, fitted parameters or detector measurements. If AI or machine learning is itself part of the research method, document the training/evaluation method and validation required for that scientific role separately from ordinary productivity use.

Turnitin checks originality, not scientific validity

Turnitin is an originality-control mechanism. A low similarity percentage does not validate a telescope calibration, Bayesian model, spectral fit, simulation, detector response or physical interpretation. Scientific validity comes from appropriate methods, transparent assumptions, uncertainty analysis, reproducibility and examiner review. Treat similarity results as one academic-integrity control rather than a quality score for the astrophysics. Likewise, automated AI indicators should not be substituted for technical evaluation of the research evidence.

Ethics, privacy and research permits

Good scientific practice applies to observational, theoretical, computational and experimental theses. Most astronomical datasets are not personal data, but a project can still include interviews, user studies, collaboration data or other information that creates privacy or ethics obligations. Not every master’s thesis automatically requires a centrally granted University research permit. Determine permit, privacy and ethics requirements from the actual project and organisational context. When personal data are involved, follow current University research-data privacy requirements and document the legal/ethical basis required for the study.

A defensible thesis chapter structure

The current public programme evidence does not impose one universal chapter template. A useful structure is Introduction, Background or Literature Review, Research Question, Data or Model, Methods, Results, Discussion, Conclusions, References and appendices where needed. Observational theses often benefit from separate Observations and Data Reduction sections. Computational theses may separate Physical Model, Numerical Method and Validation. Instrumentation theses may separate Requirements, Design, Calibration and Performance. Results should report what was measured or calculated; Discussion should interpret uncertainty, limitations and physical meaning.

UTUGradu submission, examination and publication

UTUGradu manages the electronic higher-degree thesis process, including originality checking, examination, approval, publication and archiving. Operational details can change independently of the stable 30 ECTS thesis rule, so recheck current instructions when submitting. Confirm that the intended final manuscript is uploaded, metadata are correct, any embargo or confidentiality issue is resolved, and the maturity route applicable to your background has been addressed. Do not rely on screenshots or process notes from older cohorts when current University instructions are available.

A practical Astronomy and Space Physics thesis timeline

A workable sequence is: verify the current HOPS and thesis course route; identify the research group and supervisor; define the question and evidence level; confirm data, telescope, archive, compute or instrument access; write the plan; run pilot reduction or simulations; check calibration and numerical/statistical assumptions; perform the main analysis; validate where possible; quantify uncertainty; write Methods and Results while analysis is active; complete seminar/project obligations; revise interpretation; then complete Turnitin, maturity, UTUGradu and examination steps. Build contingency time for observing failures, archive access, pipeline changes, long computations and collaboration review.

Final pre-submission checklist

Before submission, confirm that the manuscript calls the thesis 30 ECTS rather than 40 ECTS; uses the correct TÄHT7042 or FFYS7070 route; keeps MTDK1307, TÄHT7043 and project/internship studies separate; identifies raw versus calibrated versus derived data; reports model and statistical assumptions; distinguishes numerical convergence from physical validation; documents spectral, timing, image or interferometric processing; states instrument response and calibration limits; keeps prototype evidence below flight-qualification claims; preserves code/data provenance; reports uncertainty; verifies AI-assisted material; checks references and figures; and follows current Turnitin and UTUGradu instructions.