Whether you’re preparing for a science fair, applying to a summer research program, or working on a competition paper, choosing the right Astrophysics topic can make a huge difference. A topic that’s too broad can make it challenging to build a strong project. On the other hand, a well-defined topic gives you a clear research question, a manageable methodology, and something concrete to discuss with a mentor or admissions officer. When it comes to astrophysics, the range of what’s researchable at the high school level has expanded considerably. Many observatories and space agencies now release their data publicly, meaning you can work with real measurements from Gaia, TESS, LIGO, or Fermi without institutional access.
How do you pick a good astrophysics research topic?
Four factors determine whether an astrophysics research topic will actually work. First, specificity: “black holes” is not a research topic, but “estimating progenitor masses from LIGO gravitational wave strain data for confirmed binary black hole mergers” is. The narrower your question, the more defensible your methodology. Second, data and source availability. The best astrophysics topics for high school students are ones where the relevant data is already publicly archived, whether that’s spectral catalogs on VizieR, light curves on MAST, or rotation curve measurements from the THINGS survey. If you cannot access the data you need to answer your question, the project becomes much harder to complete regardless of how interesting it is. Third, personal investment. Astrophysics covers everything from solar physics to cosmology to nuclear astrophysics, and you’ll produce sharper work on a question you’re genuinely curious about. Fourth, real-world or scientific relevance: Topics connected to an active debate such as the Hubble tension or a recent discovery such as JWST’s first atmospheric spectra can place your research in a broad scientific context.
To help you get started, we have compiled this list of 25 astrophysics research topics from beginner to advanced difficulty levels. The topics cover areas such as stellar physics and galactic astronomy, gravitational wave science and astrobiology, and use both data-driven and literature-based approaches to provide a viable starting point regardless of your current math level or coding experience.
Key Takeaways
- 25 astrophysics research topics span the full range from beginner projects like tracking sunspot cycles, to advanced work like analyzing LIGO gravitational wave strain data, giving students options regardless of their current math level or coding experience.
- A good research topic depends on four factors: specificity of the research question, availability of public data, personal interest in the subject, and connection to an active scientific debate or recent discovery, such as the Hubble tension or JWST’s atmospheric spectra.
- Many of these topics rely entirely on publicly archived data, meaning students can work with real measurements from missions like Gaia, TESS, LIGO, and Fermi without needing institutional access or a university lab placement.
- Several beginner-friendly topics require no programming background at all. Classifying galaxy morphology using Galaxy Zoo data and tracking sunspot cycles using NOAA datasets can both be completed using spreadsheets like Excel or Google Sheets.
- More advanced topics, such as measuring gravitational wave signals or characterizing gamma-ray bursts, require comfort with Python and statistical methods, along with the ability to read and synthesize primary scientific literature from sources like NASA ADS.
- Several topics connect directly to major, ongoing scientific debates. The Hubble constant discrepancy, the nature of dark matter revealed through galaxy rotation curves, and the search for intermediate-mass black holes in globular clusters are all active areas where a student project can meaningfully engage with unresolved questions.
- Choosing a strong astrophysics topic is one path toward a competitive research project, but students interested in exploring research in a different subject area might consider Horizon’s Research Seminars and Labs, a mentored virtual research program with 600+ specializations across subjects like data science, political theory, and machine learning.
1. Tracking sunspot cycles using historical solar observatory data
Sub-field: Solar physics
Difficulty: Beginner
Why it’s interesting: The Sun’s 11-year activity cycle directly affects satellite operations, GPS accuracy, ppp and power grid stability, yet the mechanisms driving cycle strength variation are still debated
Suggested research question: How does sunspot count vary across recent solar cycles, and do higher-activity cycles correspond to more frequent geomagnetic storms?
Key methods/data sources: NOAA Solar Cycle Progression data, NASA’s solar event catalog, Google Sheets or Excel for plotting
Good fit for: Complete beginners comfortable with reading graphs and working with downloaded datasets
To conduct your research, you can use NOAA’s Solar Cycle Progression dataset to record daily and monthly sunspot counts going back more than a century. The core task is to plot sunspot activity across multiple cycles, and examine how the amplitude of each cycle, i.e. the number of sunspots observed at solar maximum, correlates with downstream effects such as geomagnetic storm frequency. You can also use NOAA’s solar event catalog that logs coronal mass ejections and solar flares, to compare major solar events with sunspot activity from the same period. The 11-year cycle is well established, but its strength varies considerably, and scientists still don’t have a reliable way to predict how active an upcoming cycle will be. Your research question should focus on whether higher peak sunspot count in a given cycle corresponds to more frequent or more intense geomagnetic disturbances on Earth. This project is a great fit for students interested in solar physics and astronomy, and can be completed using spreadsheets and basic data analysis techniques, without any advanced math.
2. Classifying galaxy morphology and its relationship to star formation rate
Sub-field: Extragalactic astronomy
Difficulty: Beginner
Why it’s interesting: Galaxy Zoo’s citizen science dataset revealed that galaxy shape correlates with color and age in ways that challenged earlier models of galaxy evolution
Suggested research question: Do spiral galaxies in the Galaxy Zoo dataset show systematically bluer colors than elliptical galaxies, and what does this suggest about ongoing star formation?
Key methods/data sources: Galaxy Zoo morphology dataset, SDSS color catalog, basic statistical comparison
Good fit for: Students interested in visual pattern recognition and data comparison with no programming background required
To conduct your research, you can use data from Galaxy Zoo, a citizen science project that recruits hundreds of thousands of volunteers to visually classify galaxy images from the Sloan Digital Sky Survey. The dataset links human-assigned galaxy types, such as spiral, elliptical and merger galaxies to real photometric measurements. You can use a subset of this catalog and compare color index values, which serve as a proxy for stellar population age and star formation activity across morphological classes. Blue galaxies are generally associated with active star formation, while red galaxies tend to have older stellar populations and lower star formation rates. Your research question can explore whether this relationship holds up statistically across a large sample, or whether red spirals and blue ellipticals complicate the picture. This topic doesn’t require a coding background since the Galaxy Zoo data tables can be filtered and analyzed in Excel or Google Sheets. The bulk of the research work is building your own comparison to make a statistical argument.
3. Measuring the hubble constant using galaxy recession velocities
Sub-field: Cosmology
Difficulty: Intermediate
Why it’s interesting: Measurements of the Hubble constant from the early universe and from nearby galaxies persistently disagree, a tension that may point to physics beyond the standard cosmological model
Suggested research question: Using redshift and distance data for a sample of galaxies, how closely can a student-derived Hubble constant match the accepted value, and what sources of error explain the difference?
Key methods/data sources: NASA/IPAC Extragalactic Database (NED), SDSS spectroscopic redshift data, Excel or Python for linear regression
Good fit for: Students comfortable with algebra and graphing who want to engage with a live cosmological debate
The Hubble constant describes how fast the universe is expanding, and right now two different measurement methods give answers that don’t agree with each other, a disagreement significant enough that physicists are debating whether it signals new physics. To conduct your research, you’ll approach this from the observational side by using redshift and distance data for a sample of galaxies from the NASA/IPAC Extragalactic Database. Converting redshift to recession velocity is straightforward algebra, and plotting velocity against distance gives you a line whose slope is your derived Hubble constant, which describes how fast the universe is expanding. The goal is to calculate a value for the Hubble constant and compare it with the two competing measurements currently at the center of the Hubble tension debate, approximately 67 and 73 km/s/Mpc). You can then investigate how factors such as measurement uncertainty in your own dataset might influence your result and contribute to the discrepancy. It’s a good topic to explore one of the most active debates in modern cosmology.
4. Modeling the Habitable Zone for Stars of Different Spectral Types
Sub-field: Exoplanet science / Astrobiology
Difficulty: Beginner
Why it’s interesting: Whether a planet can host liquid water depends heavily on its host star’s luminosity and temperature, and the discovery of habitable-zone planets around M dwarfs has reopened questions about what conditions life actually requires
Suggested research question: How does the habitable zone width and distance change across M, K, G, and F-type stars, and which stellar type produces the most favorable conditions for Earth-like planets?
Key methods/data sources: NASA Exoplanet Archive, published habitable zone calculator (Kopparapu et al. model), spreadsheet calculations
Good fit for: Beginners with an interest in astrobiology who want a project grounded in real stellar physics
The habitable zone is the range of orbital distances around a star where liquid water could exist on a rocky planet’s surface, and its boundaries shift substantially depending on stellar luminosity and temperature. To conduct your research, you can use the Kopparapu et al. (2013) habitable zone model, which is freely available and well documented, to calculate inner and outer habitable zone limits for stars across the M, K, G, and F spectral sequence using only their effective temperatures and luminosities. The NASA Exoplanet Archive provides both stellar parameters and confirmed planet orbital distances, so you can then check how many known planets fall within the calculated zones for their respective hosts. Your research question can focus on whether certain stellar types produce wider or more stable habitable zones than others, and factors that complicate the simple distance-based definition of habitability, such as tidal locking risk for M dwarfs, UV flux for F stars.
5. Analyzing light curves of transiting exoplanets using tess or kepler data
Sub-field: Exoplanet science
Difficulty: Intermediate
Why it’s interesting: Transit photometry was responsible for the discovery of thousands of exoplanets, and the same technique is now being used to probe atmospheric composition through transmission spectroscopy
Suggested research question: Using archived TESS light curve data, can a student accurately measure the radius ratio of a known transiting exoplanet relative to its host star, and how does this compare to published values?
Key methods/data sources: NASA MAST archive (lightkurve Python package or web interface), published exoplanet parameters on the NASA Exoplanet Archive
Good fit for: Students with some physics background who are comfortable learning basic Python or using pre-built tools
When a planet crosses in front of its star, it blocks a small fraction of the star’s light, producing a dip in the light curve that encodes the planet’s size relative to the star. To conduct your research, you can access archived light curves from NASA’s MAST portal, either using the web interface or the lightkurve Python package, and work with data for a known transiting system. From the depth of the transit dip, you can calculate the planet-to-star radius ratio and compare it to the published value on the NASA Exoplanet Archive. Factors such as limb darkening, stellar variability, and instrumental noise can affect your measurement, so part of the project involves accounting for why your result might deviate from the accepted value. If you want to extend your project, you can measure the transit duration and orbital period from multiple transits and use Kepler’s third law to estimate the planet’s orbital distance from its host star.
6. Comparing stellar populations in globular clusters using color-magnitude diagrams
Sub-field: Stellar astrophysics
Difficulty: Intermediate
Why it’s interesting: Globular clusters were long assumed to contain single-age, single-composition stellar populations, but recent observations have revealed multiple stellar generations within individual clusters, which is still not fully explained
Suggested research question: Using published photometry data for a selected globular cluster, where does the main sequence turnoff occur, and what estimated age does that imply?
Key methods/data sources: Hubble Legacy Archive, VizieR photometry catalogs, Python (matplotlib) or Excel for plotting CMD diagrams
Good fit for: Students who have covered stellar evolution in class and want to connect theory to real data
A color-magnitude diagram plots stars by luminosity against color (which maps to temperature), and in a globular cluster it takes a characteristic shape that encodes the cluster’s age. The main sequence turnoff point, where stars begin leaving the main sequence to become red giants, is the age clock: higher-mass stars evolve faster, so the position of the turnoff tells you how old the cluster is. To conduct your research, you can use photometry data for a selected globular cluster from VizieR or the Hubble Legacy Archive, plot a color-magnitude diagram (CMD) using Python or Excel, and identify the turnoff by comparing your diagram to published stellar isochrones. What makes this project particularly interesting is the question of multiple stellar populations. Several clusters, including Omega Centauri, show broadened or split sequences suggesting more than one generation of star formation, which still lacks a fully accepted explanation. You can investigate whether your chosen cluster shows similar complexity and what this might reveal about its evolutionary history.
7. Investigating the mass discrepancy in galaxy rotation curves
Sub-field: Galactic astronomy / Dark matter
Difficulty: Intermediate
Why it’s interesting: The flat rotation curves of spiral galaxies, which contradict what Newtonian gravity predicts from visible mass alone, remain one of the strongest observational arguments for dark matter
Suggested research question: For a selected nearby spiral galaxy, how large is the discrepancy between the rotation velocity predicted from visible mass and the observed rotation curve, and at what radius does the discrepancy become significant?
Key methods/data sources: Published HI rotation curve data (THINGS survey), NASA ADS for source papers, Excel or Python for calculations
Good fit for: Students interested in dark matter who want a quantitative, physics-grounded project
Newtonian gravity predicts that stars far from a galaxy’s center should orbit more slowly, the same way Neptune moves slower than Mercury around the Sun. Observed rotation curves show the opposite, that orbital velocities flatten out and stay roughly constant far into the disk. To conduct your research, you’ll work with published HI (neutral hydrogen) rotation curve data from the THINGS survey, which provides velocity measurements at different distances from the centers of nearby spiral galaxies. The calculation involves estimating the expected velocity at each radius from the visible mass distribution, then comparing it to the observed values and computing the mass discrepancy as a function of radius. Your research question should focus on where the discrepancy first becomes significant, at what fraction of the optical radius the visible mass becomes clearly insufficient, and whether this varies between high- and low-surface-brightness galaxies.
8. Variable star period analysis using aavso light curve data
Sub-field: Stellar astrophysics
Difficulty: Beginner
Why it’s interesting: Cepheid variable stars were the tool that allowed Edwin Hubble to establish that other galaxies exist and that the universe is expanding, making them a cornerstone of the entire cosmic distance ladder
Suggested research question: Using AAVSO light curve data for a selected Cepheid variable, what is the star’s pulsation period, and how does it compare to its listed period-luminosity relationship?
Key methods/data sources: AAVSO Light Curve Generator, published period-luminosity calibrations, Excel for period fitting
Good fit for: Beginners who want hands-on data work without needing advanced math
The period-luminosity relationship for Cepheid variables, discovered by Henrietta Leavitt in 1912, means that if you measure how long a Cepheid takes to brighten and dim, you can calculate its intrinsic brightness and therefore its distance. To conduct your research, you can use brightness measurements from the American Association of Variable Star Observers (AAVSO), which provides free, downloadable brightness measurements for thousands of variable stars spanning decades of observations. You’ll choose a Cepheid variable star, download its light curve, and extract the pulsation period by identifying the repeating brightness pattern, either visually or using a phase-folding technique in Excel or Python. You can then apply the published period-luminosity calibration to estimate the star’s absolute magnitude and compare it to the catalogued value. Beyond Cepheids, you can investigate whether other variable stars, such as Mira variables, show a similar period-luminosity relationship, and explore why their behavior is harder to use for distance measurements.
9. Mapping the milky way’s structure using gaia parallax data
Sub-field: Galactic astronomy
Difficulty: Intermediate
Why it’s interesting: Gaia’s third data release provided parallax and proper motion data for over a billion stars, enabling the most detailed three-dimensional map of the Milky Way ever assembled, and still contains open questions about the disk’s warp and spiral arm structure
Suggested research question: Using a subset of Gaia DR3 data, how does stellar density vary with galactic latitude, and does the distribution match the expected structure of the thin disk?
Key methods/data sources: Gaia Archive (pre-filtered subsets), ESA’s Gaia data visualization tools, Python with pandas and matplotlib
Good fit for: Students with some data comfort who are interested in our own galaxy’s structure
Gaia measures the tiny apparent shift in a star’s position as Earth orbits the Sun, and from that parallax angle it calculates the star’s distance with remarkable precision. To conduct your research, you can use data from Gaia’s third data release, which contains distances, proper motions, and photometry for over a billion stars, and a filtered subset is accessible through the Gaia Archive’s query interface. You’ll select stars within defined distance ranges and plot their spatial distribution in galactic coordinates to trace where stellar density peaks and falls off across the Milky Way. Your research question can focus on how stellar density changes with distance above or below the galactic plane, and whether the observed distribution matches the expected exponential scale height of the Milky Way’s thin disk. You can also look at how distance uncertainty propagates into your maps, which is itself a meaningful data quality exercise.
10. Pulsar timing anomalies and what they reveal about neutron star interiors
Sub-field: High-energy astrophysics / Compact objects
Difficulty: Advanced
Why it’s interesting: Pulsars are among the most precise natural clocks in the universe, but they occasionally undergo sudden spin-up events called glitches that are thought to reveal quantum mechanical behavior in their interiors
Suggested research question: What patterns in pulsar glitch frequency and magnitude have been observed across the known pulsar population, and do larger glitches correlate with specific pulsar characteristics such as spin-down rate?
Key methods/data sources: ATNF Pulsar Catalogue, published glitch databases (Jodrell Bank), NASA ADS for primary literature
Good fit for: Advanced students interested in compact object physics who are comfortable reading scientific papers
Pulsars rotate with extraordinary regularity, but they occasionally undergo glitches such as sudden jumps in spin rate followed by a slow relaxation back toward the previous trend. To conduct your research, you can use timing data from the ATNF Pulsar Catalogue and Jodrell Bank’s glitch database, both of which contain measurements and glitch events for hundreds of pulsars. Your research will involve examining whether glitch magnitude or frequency correlates with a pulsar’s spin-down rate, characteristic age, or other measurable properties. Glitches are thought to arise from interactions between the neutron star’s solid crust and a superfluid interior, but the exact mechanism and what it implies about neutron star internal structure, remains an active area of research. This project requires reading primary papers closely and evaluating how observational evidence is used to test competing scientific explanations.
11. Estimating the age of the universe from Type Ia supernova distance data
Sub-field: Cosmology
Difficulty: Advanced
Why it’s interesting: The 1998 discovery that the universe’s expansion is accelerating, which earned the 2011 Nobel Prize in Physics, came from exactly this kind of analysis, and the underlying dataset is still publicly accessible
Suggested research question: Using published distance modulus and redshift data from the Supernova Cosmology Project, can a student reproduce the evidence for accelerated expansion and estimate an age for the universe consistent with accepted values?
Key methods/data sources: Supernova Cosmology Project public data, Python or Excel for Hubble diagram construction, NASA ADS for methodology papers
Good fit for: Students with strong math skills and an interest in the observational history of cosmology
Type Ia supernovae are used as standard candles because they all reach roughly the same intrinsic brightness at peak, allowing astronomers to calculate distance from apparent brightness alone. To conduct your research, you can use the Supernova Cosmology Project, which makes its datasets publicly available, and download apparent magnitude and redshift values for a compiled sample of supernovae. Plotting distance modulus against redshift produces a Hubble diagram, and deviations from what a non-accelerating universe would predict show up as distant supernovae appearing fainter than expected. This is the actual analysis that earned Saul Perlmutter, Brian Schmidt, and Adam Riess the 2011 Nobel Prize in Physics. You can reproduce the core diagram, estimate a value for the Hubble constant, and examine whether your data independently supports an accelerating expansion or whether the effect is subtle enough to be influenced by sample size and measurement uncertainty.
12. Characterizing gamma-ray bursts by duration and spectral properties
Sub-field: High-energy astrophysics
Difficulty: Advanced
Why it’s interesting: Gamma-ray bursts are the most luminous explosions in the known universe, and while they are broadly divided into short and long categories with different physical origins, a growing number of detections do not fit cleanly into either class
Suggested research question: Using the Fermi GBM Burst Catalog, is there a statistically clear bimodal distribution in burst duration (T90), or is there a continuous spread that challenges the short/long classification?
Key methods/data sources: Fermi GBM online catalog, Python for histogram analysis and statistical testing, published classification papers via NASA ADS
Good fit for: Students interested in extreme astrophysics who want a data-heavy, statistically oriented project
The Fermi Gamma-ray Burst Monitor has detected thousands of bursts and makes its catalog freely available online. The standard classification divides bursts into short (under 2 seconds, likely from neutron star mergers) and long (over 2 seconds, likely from collapsing massive stars), based on the T90 parameter, the time over which 90% of the burst’s energy is detected. To conduct your research, you’ll download T90 values and peak energy data for a large sample of bursts, plot the distribution of burst durations, and statistically test whether the data actually supports a clean two-population model or whether the observed distribution is continuous. Several published papers have proposed the existence of a third class of gamma-ray bursts and questioned the traditional two-second dividing line, making this an active area of research. You can apply histogram analysis and basic statistical tests such as chi-squared test or a simple visual assessment of bimodality to evaluate different classification models and compare your conclusion with published research.
13. Exploring gravitational lensing as evidence for dark matter in galaxy clusters
Sub-field: Cosmology / Gravitational physics
Difficulty: Intermediate
Why it’s interesting: The Bullet Cluster is considered one of the most direct pieces of observational evidence for dark matter, because X-ray and lensing maps show the hot gas and the mass center separated during a cluster collision
Suggested research question: How do the optical, X-ray, and weak lensing mass maps of the Bullet Cluster differ spatially, and what does the offset between baryonic matter and total mass imply about the nature of dark matter?
Key methods/data sources: Chandra X-ray Observatory archive, ESO/Hubble public imaging, published lensing mass reconstructions, NASA ADS for methodology
Good fit for: Students interested in dark matter and general relativity who prefer a literature and image-analysis approach
When a massive object such as a galaxy cluster sits between us and more distant galaxies, its gravity bends the light from those background sources, distorting their shapes into arcs. The degree of distortion maps the total mass of the lens, including matter that emits no light. In the Bullet Cluster, X-ray observations from the Chandra X-ray Observatory show the hot gas – most of the ordinary matter in a cluster – lags behind the galaxy distributions after a collision. However, gravitational lensing maps show the total mass is concentrated around the galaxies rather than the gas. This spatial separation is widely regarded as one of the strongest direct evidence that most of the mass is non-baryonic. You’ll work with published mass maps and multi-wavelength images, examine the positional offset between the X-ray centroid and the lensing mass centroid, and assess what physical scenarios could reproduce this pattern without invoking dark matter.
14. Analyzing atmospheric biosignatures targeted by JWST
Sub-field: Exoplanet science / Astrobiology
Difficulty: Intermediate
Why it’s interesting: JWST’s transmission spectroscopy observations are now detecting molecular features in exoplanet atmospheres, and the scientific community is actively debating which molecular combinations would constitute convincing evidence of biological activity
Suggested research question: Based on published JWST spectral results and astrobiological literature, which atmospheric biosignature combinations are considered most robust against false positives from non-biological chemistry?
Key methods/data sources: NASA ADS, JWST Early Release Science papers, NASA Exoplanet Exploration website, Google Scholar
Good fit for: Students interested in the search for life who want a literature-synthesis project tied to active ongoing research
The James Webb Space Telescope (JWST) detects molecules in exoplanet atmospheres through transmission spectroscopy. As a planet transits its star, a small portion of starlight filters through the planet’s atmosphere, and certain molecules absorb specific wavelengths, leaving dips in the observed spectrum. The challenge for astrobiology is that molecules often associated with life such as oxygen, methane, or water vapor, can also be produced through non-biological processes. As a result, the current scientific discussion focuses on combinations of molecules, particularly disequilibrium combinations that chemistry alone wouldn’t sustain, and are difficult to explain through abiotic mechanisms alone. By reviewing JWST’s science papers on atmospheres (starting with TRAPPIST-1 planets and hot Jupiter systems), you can evaluate how astrobiologists distinguish potential biosignatures from false positives and what kind of atmospheric evidence would be required to make a compelling case for life beyond earth.
15. Simulating Orbital Dynamics Using N-Body Tools
Sub-field: Computational astrophysics / Orbital mechanics
Difficulty: Intermediate
Why it’s interesting: Most planetary systems are not two-body problems, and gravitational interactions between planets produce resonances, instabilities, and migration patterns that shaped our own solar system’s architecture
Suggested research question: Using an N-body simulator, how does adding a Jupiter-mass planet to a model planetary system affect the long-term orbital stability of smaller inner planets?
Key methods/data sources: REBOUND N-body Python package, web-based simulators like Universe Sandbox (for non-coding students), published solar system formation papers
Good fit for: Students interested in computational methods or planetary science who want to do simulation-based research
Real planetary systems involve gravitational interactions between every body simultaneously, and analytical solutions only exist for the two-body case. N-body simulations numerically integrate these interactions over time, allowing you to watch how systems evolve. To conduct your research, you can use REBOUND; this is a free Python-based N-body package designed for planetary dynamics, and includes tutorials that work for beginners. You’ll set up a test system, place a Jupiter-analog at varying orbital distances from a central star, and track whether inner terrestrial planets remain on stable orbits over simulated timescales of millions of years. Your research can focus on placing a Jupiter-like planet at different distances from a central star and examining how its gravitational pull affects the stability of inner terrestrial planets. By tracking orbital changes, you can focus on which configurations produce orbital resonances, where planets’ periods form simple integer ratios such as Jupiter and its Galilean moons, and which lead to chaotic behavior or planetary ejections. If coding is a barrier, Universe Sandbox provides a visual interface for similar experiments, though with less control over parameters.
16. Reconstructing the evidence for an accelerating universe from published supernova data
Sub-field: Cosmology
Difficulty: Advanced
Why it’s interesting: Dark energy is the leading explanation for cosmic acceleration and accounts for roughly 68% of the universe’s total energy content, yet its physical nature remains entirely unknown.
Suggested research question: Using archival Type Ia supernova apparent magnitude and redshift data, can a student demonstrate statistically that distant supernovae are dimmer than expected for a non-accelerating universe?
Key methods/data sources: Union 2.1 supernova dataset, Python or Excel for Hubble diagram, published methodology from Perlmutter et al. and Riess et al
Good fit for: Students with strong quantitative skills who want to engage directly with Nobel Prize-level science using real data
This project uses the Union 2.1 dataset from the Supernova Cosmology Project, which compiles distance moduli and redshifts for 580 supernovae with careful systematic corrections already applied. The central question is whether you can demonstrate, from the data alone, that high-redshift supernovae are systematically fainter than a matter-only expanding universe would predict. To investigate this, you’ll construct a Hubble diagram, overlay theoretical curves for different cosmological models such as matter-dominated, flat with a cosmological constant, and open universe, and assess which model best fits your plotted points. This project requires an understanding of what the distance modulus formula means and how redshift reflects the expansion history of the universe. The ultimate goal is to understand the logical chain of observational evidence well enough to explain why these 580 supernovae played a major role in convincing the astronomers that dark energy is real.
17. Measuring the properties of stars using spectral line analysis
Sub-field: Stellar astrophysics / Spectroscopy
Difficulty: Beginner
Why it’s interesting: Virtually everything we know about stars beyond the Sun comes from spectroscopy, and the positions of absorption lines encode temperature, composition, and velocity in ways that are straightforward to read once the technique is understood
Suggested research question: Using archived stellar spectra from a public database, can a student identify the spectral type and dominant absorption lines for a set of stars with known classifications, and does the line pattern match theoretical predictions?
Key methods/data sources: SDSS Spectra Viewer, ESO archive, published spectral classification atlases, Google Scholar for background
Good fit for: Beginners who want a visually intuitive introduction to how astronomers gather information about stars
Every element absorbs light at specific wavelengths, and stellar spectra are crossed with absorption lines whose positions and relative strengths reveal what the star is made of and how hot its surface is. To conduct your research, you can use the SDSS Spectra Viewer that lets you pull up optical spectra for thousands of objects and identify features by wavelength. You’ll work with a set of stars spanning the OBAFGKM classification sequence, locate key absorption lines such as the hydrogen Balmer series, calcium H and K, magnesium lines and sodium lines, and connect their relative strengths to the temperature dependence of atomic energy levels. Your research question could focus on why hydrogen Balmer lines reach their maximum strength in A-type stars rather than in the hottest O stars, a phenomenon that requires understanding of ionization physics at a qualitative level. You can also look at Doppler-shifted lines in binary star spectra and measure radial velocities, which adds a kinematic dimension to the project.
18. Examining the cosmic microwave background for evidence of Big Bang cosmology
Sub-field: Cosmology
Difficulty: Intermediate
Why it’s interesting: The CMB temperature anisotropies contain encoded information about the early universe’s density fluctuations, and the precise pattern of peaks in the power spectrum is consistent with a flat universe dominated by dark matter and dark energy
Suggested research question: How do the angular power spectrum features of the CMB measured by Planck compare to the predictions of standard Big Bang cosmology, and what would the spectrum look like under alternative cosmological models?
Key methods/data sources: ESA Planck mission public data, NASA LAMBDA data archive, Planck 2018 results papers, Google Scholar
Good fit for: Students interested in the large-scale structure and origin of the universe who are comfortable with scientific literature
The Cosmic Microwave Background (CMB) is thermal radiation left over from about 380,000 years after the Big Bang, when the universe cooled enough for hydrogen to form and photons to travel freely. Planck mission maps show temperature fluctuations at the level of one part in 100,000 across the sky, and the angular power spectrum of these fluctuations contains a series of acoustic peaks whose positions and heights constrain cosmological parameters. You’ll work with the Planck 2018 results papers and the NASA LAMBDA data archive to understand what each peak physically represents: the first peak’s angular scale determines the spatial curvature of the universe, while the ratio of odd to even peak heights provides information about the ordinary, baryon density. Your project will involve comparing the observed power spectrum to predictions from alternative cosmological models, such as models without dark matter or with different curvature, and explaining why the standard model fits and alternatives don’t.
19. Investigating how stellar mass determines a star’s life cycle and endpoint
Sub-field: Stellar astrophysics
Difficulty: Beginner
Why it’s interesting: The mass of a star at birth determines almost everything about it, from how long it lives to whether it ends as a white dwarf, neutron star, or black hole, and this relationship underpins most of what we know about stellar populations
Suggested research question: Using published stellar evolution tracks, how does the main sequence lifetime and final remnant type change across the mass range from 0.5 to 25 solar masses?
Key methods/data sources: Published stellar evolution tables, MESA stellar evolution code (for advanced students), Google Scholar, NASA ADS for review articles
Good fit for: Beginners looking for a conceptually rich, literature-based project with clear, quantifiable relationships to analyze
A star’s initial mass governs its internal temperature, fusion rate, lifespan, and eventual death, and these relationships span roughly three orders of magnitude in mass and ten orders of magnitude in luminosity across the main sequence. To conduct your research, you can use published stellar evolution models from sources such as the Geneva or PARSEC stellar models. You’ll compile main sequence lifetimes, luminosities, and remnant types of stars ranging from 0.5 – 25 solar masses and look for the quantitative relationships between them. Your research can focus on the mass-luminosity relation, which shows that luminosity increases much more rapidly than mass, causing high-mass stars to burn through their nuclear fuel thousands of times faster despite having more of it. You can explore the approximate mass range that separates stars that become white dwarfs from those that undergo core collapse and become neutron stars, and why the core collapse mechanism changes above roughly 25 solar masses to potentially produce a black hole directly without a visible supernova.
20. Comparing exoplanet system architectures around different stellar types
Sub-field: Exoplanet science
Difficulty: Intermediate
Why it’s interesting: The Kepler and TESS missions have revealed that planetary system architecture varies significantly with host star type, and understanding whether this reflects formation conditions or detection bias is an active area of research
Suggested research question: Using the NASA Exoplanet Archive, how do the median orbital period and planet radius distributions differ between confirmed planets orbiting M dwarfs versus G-type stars, and could detection bias explain the differences?
Key methods/data sources: NASA Exoplanet Archive bulk data download, Python or Excel for statistical comparison, published occurrence rate papers via NASA ADS
Good fit for: Students comfortable with basic statistics who want a data-driven project with a current research context
The NASA Exoplanet Archive’s bulk data table lets you download confirmed planet parameters for thousands of systems and filter by host star spectral type. You’ll examine key variables such as distributions of orbital period, planet radius, and multiplicity, or the number of planets within a system, between M dwarf hosts and G-type hosts. The complication is that detection bias runs in opposite directions for some of these. Transit surveys find close-in planets more easily regardless of host star type, but M dwarfs are intrinsically dimmer and closer-packed habitable zones mean that the same orbital periods correspond to different physical environments. Your project should attempt to separate the differences in the distributions that are physically real and the ones that are artifacts of how planets are found, and this requires reading occurrence rate correction papers alongside the raw catalog data.
21. Analyzing gravitational wave signals from binary compact object mergers
Sub-field: Gravitational wave astronomy
Difficulty: Advanced
Why it’s interesting: LIGO’s 2015 detection of gravitational waves confirmed a century-old prediction of general relativity and opened an entirely new observational window on the universe, with each new event revealing information about compact object populations
Suggested research question: Using LIGO open science data for a confirmed binary black hole merger, can a student identify the inspiral, merger, and ringdown phases in the strain data and estimate the total progenitor mass from the peak frequency?
Key methods/data sources: LIGO Open Science Center (GWOSC), published parameter estimation tutorials, Python with GWpy package
Good fit for: Advanced students with strong physics and math backgrounds who are willing to learn some Python
LIGO’s strain data measures spacetime distortions at the level of 10^-21 meters, smaller than a proton, making gravitational wave detections difficult to detect without data processing. To conduct your research, you’ll use data from the LIGO Open Science Center that provides strain time series data and tutorials for confirmed gravitational wave detections such as GW150914. Using the gwpy Python package, you can load the data, apply a bandpass filter to isolate the signal in the detector’s most sensitive frequency range, visualize the chirp, and see that the characteristic frequency increases as two black holes spiral inward. By measuring the frequency at merger, you can estimate the total mass of the system using the formula relating chirp mass to gravitational wave frequency. A particularly interesting research question is how different the pre-processed and raw strain data look, and what those differences reveal about the noise sources that LIGO has to overcome to identify the real signals. You can also compare multiple events from the GWTC catalog and examine whether the detected black hole population shows any mass clustering or other trends.
22. Tracing the origins of heavy elements through neutron star merger events
Sub-field: Nuclear astrophysics / Multi-messenger astronomy
Difficulty: Intermediate
Why it’s interesting: The 2017 detection of GW170817, a neutron star merger observed in both gravitational waves and light, provided direct evidence that collisions like these are responsible for producing many of the heavy elements found on Earth, including gold and platinum
Suggested research question: What is the current observational evidence that r-process nucleosynthesis in neutron star mergers accounts for the cosmic abundance pattern of elements heavier than iron?
Key methods/data sources: Published papers on GW170817 kilonova observations, NASA ADS, Google Scholar, NIST atomic spectra databases for element identification context
Good fit for: Students interested in the origin of matter who want a literature-based project connecting nuclear physics to astrophysical events
Elements heavier than iron can’t be produced in normal stellar fusion because the process requires adding neutrons faster than the nucleus can decay, a process called rapid neutron capture (r-process). Neutron star mergers were theorized as the primary r-process site for decades before GW170817 provided the first direct confirmation. The kilonova optical afterglow showed spectral signatures of freshly synthesized heavy elements, including lanthanides. To conduct your research, you can work through the published observational papers on GW170817’s kilonova and connect them to the solar abundance pattern for r-process elements, which shows characteristic peaks at specific mass numbers (around 130 and 195) that reflect neutron capture physics. Your research involves examining whether the amount of heavy elements produced in GW170817 is enough to explain the abundance of these elements in the Sun and in old metal-poor stars, and how frequently neuron star mergers must occur to account for them.
23. Investigating evidence for intermediate-mass black holes in globular clusters
Sub-field: High-energy astrophysics / Black hole physics
Difficulty: Advanced
Why it’s interesting: Stellar-mass and supermassive black holes are well established, but intermediate-mass black holes in the range of hundreds to thousands of solar masses are theoretically predicted and have recently attracted observational claims that remain contested
Suggested research question: What is the current observational evidence for or against intermediate-mass black holes in globular clusters, and why do different measurement methods yield conflicting conclusions?
Key methods/data sources: NASA ADS for primary literature, Hubble Space Telescope kinematic studies, published review papers on IMBH candidates
Good fit for: Advanced students comfortable reading contested scientific literature and synthesizing arguments from multiple observational approaches
Black holes have been well confirmed in two mass ranges: stellar-mass (roughly 5 to 100 solar masses, formed from collapsing stars) and supermassive (millions to billions of solar masses, found in galactic centers). The intermediate range, from a few hundred to a few hundred thousand solar masses, is theoretical, but observationally contested. Globular clusters are among the leading candidate environments where these objects may reside. Your research can focus on evaluating the evidence for and against IMBHs in clusters such as the Omega Centauri and NGC 1851, paying attention to the methods used to detect them, such as stellar kinematics (how fast stars move near the cluster center), X-ray and radio emission from accretion, and the behavior of tidal streams. A key research question is why kinematic measurements from different groups studying the same cluster sometimes yield contradictory conclusions, and what observational or methodological factors drive those disagreements.
24. Examining solar wind interaction with earth’s magnetosphere during geomagnetic storms
Sub-field: Solar physics / Space weather
Difficulty: Beginner
Why it’s interesting: Geomagnetic storms driven by coronal mass ejections can disable satellites, disrupt radio communications, and induce currents in power lines, making this a topic with direct technological consequences that is under active operational monitoring
Suggested research question: Do geomagnetic storm intensity (measured by the Dst index) and duration correlate with solar wind speed and particle density data recorded at the time of impact?
Key methods/data sources: NOAA Space Weather Prediction Center data, NASA OMNI solar wind database, World Data Center for Geomagnetism (Dst index), Excel or Python for correlation analysis
Good fit for: Beginners interested in space weather who want a data comparison project with direct real-world relevance
When a coronal mass ejection reaches Earth, its interaction with the magnetosphere depends on the solar wind’s speed, density, and magnetic field orientation. The NOAA Space Weather Prediction Center archives real-time and historical solar wind measurements from the DSCOVR spacecraft at the L1 Lagrange point. The World Data Center for Geomagnetism provides Dst index values, a measure of how compressed Earth’s magnetosphere becomes during a storm. Your research can focus on a set of historical storm events and examine whether solar wind speed, particle density, or southward magnetic field component (Bz) is the strongest predictor of storm intensity, as measured by peak negative Dst value. The Bz orientation matters because a southward-pointing field couples most efficiently with Earth’s northward-pointing dipole, allowing energy transfer into the magnetosphere, and your analysis can test whether that theoretical prediction holds up in the historical data.
25. Evaluating the Fermi Paradox in light of recent exoplanet discoveries
Sub-field: Astrobiology / SETI science
Difficulty: Beginner
Why it’s interesting: The confirmed existence of billions of potentially habitable exoplanets has made the absence of detected extraterrestrial signals more puzzling, and proposed resolutions to the Fermi Paradox range from rare Earth hypotheses to technological civilization self-destruction timescales
Suggested research question: Given current estimates of Earth-like planet occurrence rates from Kepler mission data, which proposed solutions to the Fermi Paradox are most consistent with both the astronomical evidence and the absence of confirmed SETI signals?
Key methods/data sources: Google Scholar, NASA ADS, Kepler occurrence rate papers, published SETI survey results, philosophical and scientific literature on the Great Filter
Good fit for: Beginners who enjoy analytical reasoning and want a project that combines astronomical data with structured scientific argumentation
The Fermi Paradox starts from the observation that a universe containing billions of potentially habitable planets should, under many assumptions, have produced detectable technological civilizations by now, and yet no confirmed signal exists. Kepler mission occurrence rate studies put the number of Earth-sized planets in habitable zones of Sun-like stars at roughly one in a few hundred, which across the galaxy translates to billions of candidates. Your research project will work through the major proposed resolutions to the Fermi paradox, such as the Great Filter, the rare Earth hypothesis, the self-destruction hypothesis, and the signaling problem, and evaluate each against the quantitative constraints from modern astronomy, such as SETI survey coverage fractions, Kepler occurrence rates, and estimates of how long technological civilizations need to persist to be detectable. This is a literature synthesis project, but the argument needs to assess which hypotheses remain consistent with what we’ve actually observed and which are increasingly challenged by the available data.
Frequently Asked Questions
Which of these topics are best for a student with no coding experience?
Tracking sunspot cycles using NOAA data, classifying galaxy morphology with Galaxy Zoo data, modeling the habitable zone for different stellar types, and analyzing variable star periods using AAVSO light curve data can all be completed using spreadsheet tools like Excel or Google Sheets, without any programming background required.
What kind of public datasets do these topics rely on?
Sources vary by topic but commonly include the NASA Exoplanet Archive, the Gaia Archive, NASA’s MAST portal for TESS and Kepler light curves, the LIGO Open Science Center, the ATNF Pulsar Catalogue, and the NASA/IPAC Extragalactic Database, most of which are freely accessible to the public.
How do I know if a research topic is too broad?
A topic like “black holes” is too broad to research, while a narrower version such as estimating progenitor masses from LIGO gravitational wave strain data for confirmed binary black hole mergers gives a defensible, specific methodology. Specificity, data availability, personal interest, and scientific relevance are the four factors that determine whether a topic will work.
Which topics connect to a currently unresolved scientific debate?
The Hubble constant topic engages directly with the Hubble tension, where early-universe and nearby-galaxy measurements disagree. The galaxy rotation curve topic connects to ongoing dark matter research, and the intermediate-mass black hole topic involves synthesizing contested observational claims that different research groups still dispute.
What if I want to pursue a research project outside of astrophysics?
These topics are focused specifically on astrophysics and related physical sciences, but students looking for a research based credential in a different subject area might consider Horizon’s Research Seminars and Labs, a selective virtual research program that pairs students with a mentor to develop an original research paper across subjects like data science, political theory, and machine learning.
Image source: Horizon Academic Research Program