Linn Boldt-Christmas

@nplinnspace.bsky.social

3787 Followers 426 Following

🇾đŸ‡Ș🚀đŸ‡ȘđŸ‡ș Astrophysicist in exoplanet science & space policy research đŸȘđŸ”­ Also: astrobiology, telescopes, climate, equality/EDI, and scicomm ✹ (she/her) – linnboldtchristmas.wordpress.com

Cerro Panchón, Chile, after a snow storm — gorgeous white snow is everywhere!!
Don’t throw away your rubbish ground-based transit data (yet)! -- A cautionary tale starring warm super-puff WASP-107 b and VLT The problem: Ground-based transmission spectroscopy observations have seen remarkable success in characterising hot Jupiter exoplanets in the last decade with increasingly high detail, and our hope is to eventually characterise smaller, cooler exoplanets similarly – one day, even Earth-like targets. To get there, we have to explore what shortcomings exist in our current observational capabilities and methodology by exploring an ”in-between” target: the neither-temperate-nor-hot Neptune-sized WASP 107-b (Teq ≈ 740K). Method: WASP-107 b is already well-studied from space, and thus a good playground for assessing the quality of our ground-based approaches. It is observationally favourable in some ways (e.g. good scale height) but more challenging in others (e.g. cloudy, cooler Teq), making it a good case study for our tried-and-tested techniques like cross-correlation. A number of molecular species had already been detected by JWST, but could we do it from the ground? By using a combination of real and simulated data, we could explore what aspects of traditional methods perform well for these challenging targets, and which need a rethink. We found that for cooler and cloudier targets like WASP-107 b, the amount of signal that one can reasonably expect is unfortunately very low. Several factors can compound and effectively create scenarios where there is simply not enough signal to reach detections of S/N ≄ 5, even if everything is done ”correctly”. What looks like bad data might not be bad, just
 as good as we can get, for now!
A pathetic attempt by AI sloppery to make a HR diagram, generated by one of our BSc student, and shared here with their permission. Honestly, I can't even begin to explain all the deviations. Imagine a normal HR diagram but one that appears to you in a fever dream. It gets so much worse the longer you look at it. WHY IS THIS ALLOWED TO EXIST AAAAA
Final paragraph of the paper: As the combination of new instrumentation and maturing methodology brings research further into the realm of characterising smaller and cooler exoplanets with high-resolution spectroscopy, the parameter space of possible exoplanetary atmospheres expands – and with it, the parameter space of uncertainties. This diversification marks exciting new scientific frontiers, but associated with this reduced error budget are many sources of noise and uncertainty that graduate from being negligible to impactful. Only by mapping out these pitfalls may we avoid them, and only then will we continue our progression towards characterising even cooler and smaller exoplanet targets with high-resolution transmission spectroscopy.
Fig. 11. from the paper. Caption: Real and simulated Kp − v plots for SYSREM iterations 3-10, cross-correlating with the H2O template including clouds. In each plot, the white dotted line denotes the expected location of the peak, and the yellow dashed line denotes the actual location of the peak. First row: the real data Kp − v map as shown in Fig. 10. Second row: a simulated observation including all expected components (a “full” simulation). Third row: a simulated observation with an intentionally perturbed wavelength solution (“bad λ”). Fourth row: a simulated observation excluding telluric contamination (“no T”). Fifth row: a simulated observation excluding a planetary atmosphere (“no P”).
Fig. 3 from the paper. Caption: Per-phase differences (black lines) between observed WASP-107 b spectra and an inverse model of the mean nightly telluric absorption with mean PSF profile scaled to appropriate airmass. This spectral segment is totally dominated by telluric features and contains no significant stellar lines. Vertical offsets between each subsequent phase were added for visibility. A blue overplotted spectrum shows the mean telluric spectrum for each night at the lowest airmass. Note that substantial structure in the residuals is only visible for two phases on Night 2 (third from bottom, and topmost), which is indicative of the PSF changes in those phases, but also to the general stability of the PSF otherwise.
Fig 5. from the paper. Caption: Simulated Kp − v plots to test the impact of including or excluding cloud decks in four different test cases. Top: Simulations of a “clear” WASP-107 b that has no cloud deck present, cross-correlated with a “clear” template that excludes clouds (left) and with a “cloudy” template that includes clouds (right). Middle: Same test as the top row, now simulating a “cloudy” WASP-107 b with a cloud deck present at P = 10−5 bar, shown on the same colour scale as the test above for comparison. Bottom: Same test as the middle row, now shown on its own colour scale for reference.
Poster entitled ”HARPS3 in commissioning” by William Brilliant from the University of Cambridge on behalf of the HARPS3 consortium
Figure caption: (a) Dorsal skin luminance in eastern tree frog (Hyla orientalis) males inhabiting across a gradient of radiation inside (CEZ) or outside the Chornobyl exclusion zone (outside CEZ). (b) Range of dorsal skin luminance in H. orientalis males (from left to right: Luminance values of 5, 20, 30, 40 and 60).
A page of detailed, cartoony schematics drawing and explaining how the wind interacts in a binary Wolf-Rayet star system, part 1. You should click the link below to check it out!
Screenshot from an English translation of the following website: https://www.uu.se/institution/fysik-och-astronomi/forskning/astronomi-och-rymdfysik/nyheter/arkiv/2025-10-31-pris-for-samverkan-och-utatriktat-arbete-till-linn-boldt-christmas The title says "Award for collaboration and outreach to Linn Boldt-Christmas" and stands above a picture that shows me against a blackboard, holding flowers and a medal. Photo credit: Camilla Thulin.
”The Far Side” comic from 1987 by Gary Larson. Two chimpanzees are in a tree, and one (seemingly female) is picking hair off another (seemingly male). She says: ”Well, well — another blond hair
 conducing a little more ’research’ with that Jane Goodall tramp?”
Figure caption: Figure 2. The hydrogen atom. The ten innermost orbits possible for the single electron of the atom of hydrogen are diagrammatically represented. All possible quantum transitions are included as follows: — short dashes, Lyman series, terminating at a 1-quantum orbit; full lines, Balmer series, terminating at a 2-quantum orbit; long dashes, Paschen series, terminating at a 3-quantum orbit. Transfers are only possible between orbits with azimuthal quantum numbers differing by ±1.
An image showing four types of exoplanets as a % of the 5000 exoplanets found. Image credit: NASA/JPL-Caltech Taken from https://science.nasa.gov/universe/exoplanets/cosmic-milestone-nasa-confirms-5000-exoplanets/ Caption: The more than 5,000 exoplanets confirmed in our galaxy so far include a variety of types – some that are similar to planets in our solar system, others vastly different. Among these are a mysterious variety known as “super-Earths” because they are larger than our world and possibly rocky.
Text from the paper: 5. DISCUSSION AND CONCLUSIONS We have presented a novel method for recovering the high-resolution transmission spectrum (effective blocking area of a planet) from high-resolution transmission spectroscopy observations. The new method is formulated as an inverse problem and it reconstructs three main functions: stellar spectrum, telluric spectrum, and the planetary transmission spectrum. The first two can be verified independently against the observations outside transit or observations of a telluric standard, where they show good agreement with expectations. The main benefit of using TSD is that it does not include any model assumptions about the exoplanetary atmosphere, its chemical composition, pressure-temperature gradients, or vertical stratification. This enables model-based analyses to be done on the resulting transmission spectrum, not unlike classical stellar spectroscopy. TSD can naturally combine multiple transits as it in fact requires more than one transit to work. Ideally, several transits (5-7) with strategically selected barycentric velocities should be combined to obtain the best results, as this enables us to recover planetary features that are “behind” telluric absorption on a given night. Many exoplanets are known to have supersonic winds, and if these are globally coherent, TSD will show their signatures in the recovered line profiles.
A copy of my PhD thesis!!!!! It is a white book with the title near the top of the cover page ”Challenges in the characterisation of exoplanetary atmospheres” The author name, Linn Boldt-Christmas, is on the bottom left of the cover page, and the logo and title of Uppsala University is on the bottom right. It is being held up by me to the window, with Ångström Laboratory visible in the background.
Slide title: Conclusions We use VLT/CRIRES+ data to study WASP-107 b at ~2.0–2.5 ÎŒm We detect CO and H2O ~6σ and ~4.5σ respectively We also achieve ~6σ using the global spectrum of Welbanks+2024 Our results are sensitive to cloud inclusion, more so than VMR These results are crucially supported by previous results and simulations and would not have been reached without them This is a smaller, cooler planet than UHJ/HJ and we underline a clearly decreased error budget for targets like these moving forward
Slide title: Is that an off-set
? Maximum of Kp-vsys map is located above expected location? We explore causes in paper, but at least partially‹due to tellurics (worst for H2O) and/or SYSREM‹as simulations show this‹”chimney smoke” also Not necessarily an offset, but an additional region of higher S/N located above true signal
Slide title: Which template(s) fared best? All three cases (CO, H2O, ”Global”) have higher detections significances with templates that include clouds The differences in detection significances is more dependent on cloud inclusion than on retrieval parameters (VMR) This is true even for e.g. CO which has a VMR difference of ~1 dex between the two. ‹Cloud-metallicity degeneracy
?
The Kp-vsys maps of the four template combinations, giving CO ~ 6.3 S/N; H2O ~4.5 S/N; "global" ~ 6.3 S/N.
Four example templates from the two retrieval frameworks, Aurora and CHIMERA, with and without clouds.
Fig. 4 from my paper. The plots show simulated Kp − vsys plots to test the impact of including or excluding cloud decks in four different test cases. Top: Simulations of a “clear” WASP-107 b that has no cloud deck present, cross-correlated with a “clear” template that excludes clouds (left) and with a “cloudy” template that includes clouds (right). Middle: Same test as the top row, now simulating a “cloudy” WASP-107 b with a cloud deck present at P = 10^−5 bar, shown on the same colour scale as the test above for comparison. Bottom: Same test as the middle row, now shown on its own colour scale for reference
Fig. 3 from my paper. It is a plot showing the simulated global (i.e. including all species) transmission spectrum of WASP-107 b, coloured by species contribution, based on parameters from the results of the Welbanks et al. (2024) CHIMERA retrieval and as generated by petitRADTRANS at a resolving power of R = 10^6. The plot demonstrates the notable change of the spectrum’s shape when either excluding (top plot, showing a “clear” atmosphere) or including (bottom plot, a “cloudy” atmosphere) the presence of a grey cloud deck at P = 10−5 bar. The grey areas denote wavelength ranges that fall outside of the spectral orders of the VLT/CRIRES+ K-band of ∌2.0–2.5 ”m (1972–2452 nm).
A screenshot from my thesis! Figure caption: A plot from Gandhi et al. (2020) [https://academic.oup.com/mnras/article/498/1/194/5899070] simulating the impact of cloud decks at different pressures (altitudes) on transmission spectra of the 1.4 ”m water feature for exoplanet GJ 436 b, across a wider wavelength range (left) versus a shorter range (right). The transmission spectra is shown at high spectral resolution, and the diamond markers indicate the equivalent HST WFC3 spectra. The height of the H2O spectral feature is given in parts per million (ppm).
A figure with the same headline "Bright, central peak = detection (in simple terms)" for Night 1, Night 2, and combined 1+2. They look like dogshit (am I allowed to swear in alt text? Does anyone read these?) and below it says "What happened?!" and followed by "...This is why this is a *methodology* paper!"
An example of a Kp-vsys map from Fig 3 of Hoeijmakers et al. 2018 (Nature volume 560, pages 453–455). The headline says "Bright, central peak = Detection (in simple terms)" to summarise what these Kp-vsys plots effectively try to show. It shows the plots for four different species.
A really dumb and poorly photoshopped picture of a Hannah Montana poster, showing Miley Cyrus as both Hannah Montana (popstar) and Miley whatever-her-name-was-in-the-show (normal girl?). They have JWST and VLT as their heads respectively, and the poster slogan of "best of both worlds" is slapped across them. Trust me, no one needs to see this.
A screenshot of this Nature website: https://www.nature.com/nature/volumes/630/issues/8018 The website says: Volume 630 Issue 8018, 27 June 2024 Popcorn planet The cover shows an artist’s impression of the exoplanet WASP-107b, a Neptune-sized world some 200 light-years (61 parsecs) from Earth in the constellation Virgo. The planet has posed a problem for researchers because although its mass is similar to that of Neptune, its density is much lower, even less than Jupiter or Saturn. In this week’s issue, Luis Welbanks and colleagues suggest an explanation for how such a small planetary core can have such a large gaseous envelope. To investigate this mystery, the team used infrared spectroscopic data from the NIRCam and MIRI instruments on JWST as well as the WFC3 instrument on the Hubble Space Telescope. The results show that WASP-107b has a high internal temperature, which suggests that its eccentric orbit around its star causes tidal-driven heating that puffs up its atmosphere. Cover image: Adina Feinstein The cover image is an illustration of WASP-107 b, which is shown as a hazy, Jupiter-like gas giant planet being illuminated by a white-ish host star in the background.
Screenshot of the home page, Colblindor, of the link in my post. The text says: "Coblis — Color Blindness Simulator If you are not suffering from a color vision deficiency it is very hard to imagine how it looks like to be colorblind. The Color BLIndness Simulator can close this gap for you. Just play around with it and get a feeling of how it is to have a color vision handicap. As all the calculations are made on your local machine, no images are uploaded to the server. Therefore you can use images as big as you like, there are no restrictions. Be aware, there are some issues for the “Lens feature” on Edge and Internet Explorer. All others should support everything just fine. So go ahead, choose an image through the upload functionality or just drag and drop your image in the center of our Color BLIndness Simulator. It is also possible to zoom and move your images around using your mouse – try it out, I hope you like it."
A close up of the two images’ respective capture of a small region of the night sky, taken from the images in the quoted post. The new Rubin image has significantly more stars.
Image of text reading: For the past decade, social media users have been like restaurant diners who can only eat at one restaurant, where they can’t see the menu in advance, the chef changes the recipes based on his mood, and the only thing diners can do if they don’t like the menu is yell loudly and hope the chef makes something different. Bluesky is more like a food court where you can choose from multiple vendors, see what each one offers, and even set up your own stand if you want. Some people still yell loudly, but out of the learned habit that that’s the only thing you can do.
An image of my title slide entitled ”Obstacles in observing exoplanetary atmospheres”. Subtitle: Pre-defence seminar, Linn Boldt-Christmas, Uppsala University, 23 May 2025. The UU symbol is on the bottom right and the image shows an artist illustration of an exoplanet and its moon around a red dwarf star.
That xkcd comic / meme where a big tower of blocks is being held up by just one single block. In this case, the block is “academics” and what it is holding up is “the backpack industry”
Average weather for Uppsala, Sweden: July is the hottest month in Uppsala with an average temperature of 16°C (61°F) and the coldest is February at -3°C (27°F) with the most daily sunshine hours at 11 in June. The wettest month is August with an average of 90mm of rain.
Plot showing recent monthly mean CO2 at Mauna Loa Observatory. 2020 starts at ~413 ppm (CO2 mole fraction) and 2025 reached ~427 ppm.
Screenshot of Principal Skinner of The Simpsons in two panels: (1) "Am I out of touch?", he asks himself; (2) "No, it's the children who are wrong", he reassures himself.
Screenshot of the 1992 Nature paper by Wolszczan & Frail on ”A planetary system around the millisecond pulsar PSR1257+12” i.e. the study broadly considered to be the first definitive detection of an exoplanet, published 9 January 1992
A photograph showing the FyrisÄn river that flows through Uppsala, taken in the city centre with a view of the Cathedral in the background
Screenshot from the Wikipedia page showing the The Orbiting Frog Otolith (OFO) spacecraft. It is a black-and-white photograph of a contraption that is dome shaped, striped, and has some panels on the bottom. I think there is a frog inside it?!
Meme of Ralph Wiggum on the bus saying ”(chuckles) I’m in danger”
 which is wHAT I AM BECAUSE THIS PAPER IS TAKING SO LONG AUUGHHH
Scanned pages 96-97 of ”Hjelpreda I HushĂ„llningen För Unga Fruentimber” (1755) by Anna Christina ”Cajsa” Warg, with mention of ”Kiöttbullar” (köttbullar ie meatballs)
Screenshot from Google Images of Swedish ”fiskbullar” dish ie balls made of fish, much like meatballs, in a sauce with potatoes and dill
A screenshot of the "Pathfinder" interface. In the "Ask me anything" search window, the question is "does the chemical composition of an exoplanet atmosphere depend on the planet's mass?". The number of papers to be retrieved has been set to 10.
Group photo from the 2024 IAF IAC IPMC YP Workshop Final Event which was held in Milan, Italy, in October 2024. The group is of ~50 people in business smart clothes and most are wearing lanyards of the events.
Schematic showing that ingress = as the planet enters onto the face of the star & egress = as it exits
Figure 7 from the paper, demonstrating notable variations in how these results manifest in a single "night"
Flow chart Start: Instrument and system parameters? A: Low R / Long td B: High R / Short td C: High R / Long td D: Low R / Short td E: Mid R / Mid td A -> Long exp B -> Short exp C, D, E -> Signal strength? Signal strength high -> short exp Signal strength low -> long exp
Exposure triangle, showing how one can increase brightness by adjusting shutter speed, or aperture, or ISO.
Figure 5 for paper, demonstrating a heat map for the four different cases where Time resolution is on the y-axis and SYSREM is on the x-axis
Picture of beautiful VLT telescope at Cerro Paranal, Chile, at sunset
Summary of conundrum: Time optimisation problem Exposures must be taken across transit to capture RV information. Either: fewer longer exposures Pros: High light collection = high signal-to-noise ratio Cons: Wide movement/exposure = high smearing Or: several shorter exposures Pros: narrow movement/exposure = low smearing Cons: low light collection = low signal-to-noise ratio Where is the compromise? Does one exist? Not too smeared vs not too noisy?
Animation showing the relative movement of the exoplanet across a single exposure. In a long exposure, the planet moves a lot, which creates more smearing. In a short exposure, the planet moves less, which creates less smearing.
Animation showing how each instance of an overhead (shown by a pink rectangle) will build up and eat into your observing time. Shorter exposures = more overheads = less light collection. Nooo!
Animation showing that if you have many exposures, you have many gaps between each exposure, where your instrument will need some time (i.e. an "overhead"). With more instances of being "between" an exposure, you will have more instances of needing time for overheads.
Animation showing how one can either take lots of short exposures or several longer exposures across the transit duration. This combination of X exposures of Y length is called the time resolution.
Animation showing a simplification of how we can capture this movement by taking exposure #1 when the lines are very blue, exposure #2 when the lines are "in the middle", and exposure #3 when the lines are very red.
Animation demonstrating how the planetary lines move across the transit from blue to red, compared to the stellar and telluric lines which remain still.
The exoplanet is in orbit around its host star. As it moves (from, say, let to right) it will be blueshifted as it enters the transit from the left and redshifted as it exists the transit to the right. These lines will have a high radial velocity shift, but comparatively, the stellar lines and the telluric lines will not.
Schematic showing how a transmission spectrum builds up: the light from a host star has its own stellar spectral lines; then lines from the exoplanet atmospheres get added to the spectrum as the starlight passes through; and finally, everything gets ruined by Earth's atmosphere right at the end which imprints its own spectral lines. How can we tell which lines are from the star and from the Earth? How can we tell which are from the planet?