Verify an Exoplanet Transit in Public Telescope Data
This guide walks you through verifying a possible exoplanet transit using publicly available telescope data, such as from the Kepler/K2 or TESS missions. You will learn how to download light curve data, identify the transit signal, and check whether it is a genuine planet signature rather than a false positive. No prior coding experience is required, though basic familiarity with spreadsheets or Python will help.
What You’ll Need
- A computer with internet access.
- A web browser (Chrome, Firefox, or similar).
- A spreadsheet program (e.g., Excel, Google Sheets) or a Python environment (e.g., Jupyter Notebook) for data analysis.
- Optional: A free account at the Mikulski Archive for Space Telescopes (MAST) to download data.
Step-by-Step
- Identify the target star and its known transit parameters. If you have a specific star in mind, note its name (e.g., HD 209458) or coordinates. For a candidate from a citizen science project, you will have the target ID and the expected transit period and duration. If you are browsing, choose a star with a confirmed or candidate planet from the NASA Exoplanet Archive.
- Go to the MAST data portal. Open mast.stsci.edu in your browser. This is the official archive for TESS, Kepler, and K2 data. In the search box, enter your target’s name or coordinates.
- Select the appropriate mission and data product. For TESS, choose the ‘TESS’ mission and then ‘Light Curves’. For Kepler/K2, choose ‘Kepler’ and then ‘Light Curves’. You will see a list of available files. Look for a file with a name like ‘hlsp’ (high-level science product) or the mission-specific light curve (e.g., ‘ktwo’ for K2).
- Download the light curve file. Click on the file name to open its details, then download the FITS file or a CSV/TSV version if available. For simplicity, prefer a ‘Simple Aperture Photometry’ (SAP) or ‘Pre-search Data Conditioning’ (PDC) light curve, which is already cleaned of many instrumental effects.
- Open the data in your analysis tool. If using a spreadsheet, import the CSV/TSV file. If using Python, you can use the
astropyandlightkurvelibraries to read the FITS file. The light curve will have columns for time (in days, often as BJD or TJD) and flux (in electrons per second or normalized). - Plot the light curve. Create a scatter plot of flux (y-axis) versus time (x-axis). You should see a mostly flat line with some scatter. If the star is variable, you might see a periodic wave. The transit will appear as a periodic dip in flux.
- Phase-fold the data. If you know the orbital period (P) from the candidate’s parameters, you can fold the light curve. In a spreadsheet, create a new column for ‘phase’ using the formula:
((time - t0) / P) - floor((time - t0) / P), where t0 is a reference transit time (mid-transit). In Python, use thelightkurvefold()method. After folding, the transits should align at phase 0. - Inspect the folded light curve. Look for a clear, repeated dip that is consistent in depth and duration. The transit depth is the fractional decrease in flux (e.g., 1% for a Jupiter-sized planet around a Sun-like star). The duration is typically a few hours. A real transit has a ‘U’ shape (or flat-bottomed ‘U’ for larger planets) that is periodic.
- Check for secondary eclipses and variations. A genuine transit should show only a primary eclipse (planet passing in front). A secondary eclipse (planet passing behind the star) may be visible for hot Jupiters but is usually much shallower. Also, check that the transit depth is consistent across multiple events; if the depth varies significantly, it may be a false positive.
- Rule out common false positives. The most common mimics are eclipsing binary stars (where the dip is deeper and often V-shaped) and blended eclipsing binaries (where the light is diluted by a companion star). To check, look at the target’s pixel data (via the ‘Target Pixel File’ on MAST) to see if the dip occurs at the same location as the star. Also, check the centroid shift: a real transit does not shift the star’s position, but a background eclipsing binary might.
- Compare with known exoplanet databases. Use the NASA Exoplanet Archive or the Exoplanet Follow-up Observing Program (ExoFOP) to see if the transit parameters match a known planet. If your candidate is new, you can submit it to the Exoplanet Watch project for further analysis.
Common Mistakes
- Using the wrong time system. Kepler and TESS data use BJD (Barycentric Julian Date) or TJD (TESS Julian Date). If you mix these with Earth-based UTC times, your phase folding will be off.
- Not removing long-term trends. Stellar variability or instrumental drift can mask or distort the transit. Always detrend the light curve (e.g., by dividing by a moving average) before folding.
- Ignoring the difference between SAP and PDC light curves. SAP light curves contain more systematic errors; PDC is corrected for many known effects. Prefer PDC for analysis.
- Overinterpreting a single dip. A single dip could be a cosmic ray hit or a data glitch. Always require at least two or three transits at the same period and depth.
- Assuming the transit is real without checking the target’s pixel data. A background eclipsing binary can cause a dip that looks like a transit. Always check the centroid and the pixel response.
Tips
- Use the
lightkurvePython library—it simplifies downloading, flattening, and folding light curves dramatically. - Start with a known exoplanet (e.g., HD 209458b, WASP-19b) to practice; you can find their data on MAST and verify the transit easily.
- When folding, use a period that is accurate to at least 6 decimal places; otherwise, the transits will smear out.
- Check the transit depth in physical units: depth = (planet radius / star radius)^2. For a Jupiter-radius planet around a solar-radius star, depth is about 1%.
- If you are using a spreadsheet, use the ‘LINEST’ or ‘SLOPE’ functions to detrend linear trends before folding.
Troubleshooting
- No transit is visible after folding. Double-check your period and t0. If the period is off by even 0.1%, the transits will be smeared. Try a period search using the Lomb-Scargle periodogram (available in
lightkurveorastropy). - The light curve is very noisy. Ensure you are using the PDC light curve, and try a larger aperture (if using TESS, you can download the ‘aperture’ file and use a different mask). Also, exclude data points taken during spacecraft thruster firings (often flagged in the quality column).
- The transit depth is too shallow or too deep. Check if the star is a binary or has a companion that dilutes the light. You can look up the star’s spectral type and radius to estimate the expected depth.
- You see a dip but it is not periodic. It could be a stellar flare or a data artifact. Check the time series for other dips at the same phase; if none, it is likely not a transit.
- Data download fails. Try using the MAST API directly or use the
lightkurvesearch function, which can download data in one line of code.
Frequently Asked Questions
What is the difference between a transit and an eclipse?
A transit is when a smaller body (planet) passes in front of a larger body (star), blocking a tiny fraction of light. An eclipse is when a larger body blocks a smaller one, like the Moon eclipsing the Sun. In exoplanet terms, we usually talk about transits for planets, and secondary eclipses when the planet passes behind the star.
How long does a typical transit last?
For a planet in a close orbit (period of a few days), the transit lasts about 1-3 hours. For planets in wider orbits (like Earth’s year), the transit can last up to 16 hours. The duration depends on the star’s size, the planet’s orbital speed, and the impact parameter.
Can I use data from ground-based telescopes?
Yes, but ground-based data are affected by atmospheric extinction and weather. Public datasets like those from the Kilodegree Extremely Little Telescope (KELT) or the SuperWASP project are available, but they are more challenging to analyze. Space-based data (Kepler, TESS) are cleaner for beginners.
What is the best way to learn the analysis?
Start with the TESS tutorial or the lightkurve documentation. They provide step-by-step examples using real data. Also, consider joining the Exoplanet Watch program, which offers guidance and a community of amateur astronomers.
How can I tell if the transit is caused by a planet or a star?
Check the depth and shape. A planet transit is typically less than 2% deep (for a Jupiter-sized planet) and has a flat bottom. An eclipsing binary star can have a depth of 10% or more, and the shape is often V-shaped because the stars are similar in size. Also, if you see a secondary eclipse (a second dip at phase 0.5), it is likely a binary star system, not a planet.
What should I do if I think I found a new exoplanet?
First, double-check your analysis and rule out false positives. Then, report your candidate to the Exoplanet Watch or the NASA Exoplanet Archive as a ‘candidate’. Do not claim a discovery without follow-up observations from professional observatories. Amateur discoveries are rare but have happened; proper vetting is essential.
Related
Method: the standard formula for this calculation, rounded as each result states. See how the calculators work or report an error.