TIC-65910228 b / NGTS-38 b, a 180 day transiting warm super-Jupiter 2026 TIC-65910228 b / NGTS-38 b, a 180 day transiting warm super-Jupiter – 8 Toby Rodel , Solène Ulmer-Moll , Samuel Gill , Christopher. A. Watson , Yoshi Nike Emilia Eschen , Alix V. Freckelton , Annelies Mortier , Karen A. Collins , Diana Dragomir , Zahra Essack , Brett Skinner , Niamh Mallaghan , Peter J. Wheatley , David R. Anderson , Ioannis Apergis , Khalid Barkaoui , Matthew P. Battley , Daniel Bayliss , François Bouchy , Edward M. Bryant , Matthew R. Burleigh , Benjamin M. J. Cadell , Samuel J. Carlier , Yann Carteret , Sarah L. Casewell , Alastair B. Claringbold , Jean C. Costes, Benjamin D. R. Davies , Lauren Doyle , Phil Evans , Jorge Fernández Fernández , Emile Fontanet , Edward Gillen , Michael R. Goad , George Harvey , Faith Hawthorn , Katlyn L. Hobbs , Melissa Hobson , Giovanni Isopi , James S. Jenkins , Alicia Kendall , David Kipping , Monika Lendl , Franco Mallia, Christopher Mann , James McCormac , Ernst J.W. de Mooij , Maximiliano Moyano , Arianna Nigioni , Mohammad Odeh , Vera Maria Passegger , Suman Saha , Richard P. Schwarz , Amber Sedgley , Avi Shporer , Abderahmane Soubkiou , Stéphane Udry , Dimitri Veras , Jean. P. Vignes, Steven Villanueva Jr. , José I. Vinés , Richard West , Thomas G. Wilson , Hannah L. Worters, Mitchell E. Young , Aldo Zapparata , Authors’ institutions are listed at the end of the paper Abstract We present the discovery of TIC-65910228 b / NGTS-38 b, a giant exoplanet with a radius of 1.081 ± 0.047 1.081\pm 0.047 R J and a mass of 4.78 − 0.37 + 0.40 4.78_{-0.37}^{+0.40} M J on a long-period ( 180.52791 ± 0.00038 180.52791\pm 0.00038 day), moderately eccentric ( e = 0.308 ± 0.011 0.308\pm 0.011 ) orbit transiting a bright (V= 10.230 ± 0.020 10.230\pm 0.020 mag) metal rich ([Fe/H]= 0.33 ± 0.09 0.33\pm 0.09 ‘dex’) F6V-F7V type host star. The planet was initially detected from a single transit in TESS Sector 33. A photometric monitoring campaign of 228 nights with NGTS detected a transit egress of the planet, which together with spectroscopic radial velocity monitoring with CORALIE and HARPS identified an orbital period of 180.5 d. These radial velocity measurements also showed the mass of the companion to be planetary. Additional transit observations coordinated by the TESS follow-up observing program allowed further confirmation and refinement of this period. With its relatively cool equilibrium temperature of 458 ± 11 458\pm 11 K, NGTS-38 b joins a small but growing population of well characterised transiting warm-Jupiters and has one of the longest periods of any discovered to date. The target is situated in the LOPS2 field of the upcoming PLATO mission which will allow for greater refinement of the system parameters and potential for the discovery of additional companions too small and/or too long-period to be seen by TESS or NGTS. NGTS-38 b’s bright host star and wide orbital separation make it an attractive target for further study, including potential measurement of its spin-orbit alignment or targeted exomoon/ring searches. 1 Introduction Despite the large number of confirmed transiting planets ( 2 ; 35 , 4520 listed on the NASA Exoplanet Archive;) 1 1 1 Accessed 2026 January 16. Available at: https://exoplanetarchive.ipac.caltech.edu , periods on the order of tens to hundreds of days remain observationally rare. A relative few (\lessapprox 25\%) have periods beyond 25 days and even fewer (\lessapprox 5\%) have periods >100 days. Many of these P\geq 100 days period planets (\sim 90\%) were discovered by the Kepler mission due to its long observational baseline. However, Kepler ’s relatively narrow field of view meant that many of these targets orbit faint, distant stars. This limits the extent to which they can be characterised in detail through spectroscopic follow-up observations. Many of the most pressing questions in exoplanet science require a large sample of well characterised long-period ( \sim 10-1000 days) transiting planets with fully measured masses, radii and orbital solutions. For instance, the existence of close in giant planets, often termed ‘hot-Jupiters’ ( 116 , e.g. 51 Pegasi b;) , challenged pre-existing notions of planet formation and evolution ( 45 , see review;) . The existence of these planets can either be explained with in-situ formation ( 25 , e.g;) or ex-situ formation followed by inward migration with the latter usually being preferred. Two main migration channels are proposed: high eccentricity migration ( 109 ; 106 ; 138 ) where a dynamical interaction perturbs the planet into an eccentric orbit that is circularised through tidal interactions over time; or disk migration ( 69 ) where interactions with a protoplanetary disk cause a planet to lose angular momentum and migrate inwards. These migration mechanisms should leave detectable dynamical traces in a planetary system, for instance, the orbital eccentricity and/or the spin-orbit alignment / obliquity between the planet and host star ( 123 ; 45 ; 4 ) . However, for giant planets with orbital periods of 10 days or shorter, tidal interactions with the star circularise orbits and can dampen obliquities ( 5 ; 4 ) - effectively removing any trace of the system’s dynamical history. Longer-period giant planets (‘warm-Jupiters’) transiting nearby bright stars are therefore valuable tracers of planet formation and migration as spectroscopic radial velocity observations can constrain their undamped orbital eccentricities and obliquities via the Rossiter-McLaughlin effect (RM; 146 ; 118 ; also see Review by 164 ). Already, RM surveys have found a greater trend towards alignment in giant planets with periods on the order of 10s to 100s of days compared to those with periods shorter than 10 days ( 141 ; 170 ; 56 ) . Although the numbers are currently small, this could suggest warm-Jupiters may be a distinct sub-population rather than representing a transitional stage towards becoming hot-Jupiters. Longer-period planets also generally experience less irradiation from their host stars. This drastically reduces the rates of both atmospheric mass loss ( 149 ) and inflation ( 45 ) , meaning these planets are more likely to retain their primordial atmospheres and interiors. If the host star is bright, the planetary atmosphere can be probed via transmission spectroscopy ( 32 , e.g;) . These observations can provide additional constraints on planet formation and migration mechanisms using tracers such as the C/O ratio ( 126 ) , as well as general atmospheric dynamics and compositions at cooler equilibrium temperatures ( 59 ) . Long period giant planets are also valuable targets in searches for circumplanetary objects such as exomoons and exorings ( 160 ) , both of which are expected to become more stable with increasing orbital separation ( 13 ; 12 ) . In addition, as we push to longer-periods, wider separations and cooler equilibrium temperatures, the possibility of either directly habitable exoplanets or habitable moons around giant planets becomes greater. The Transiting Exoplanet Survey Satellite ( 142 , TESS ;) is specifically designed for the discovery of transiting exoplanets around nearby bright stars. The survey strategy of tiling TESS ’ wide rectangular field across the sky in ‘sectors’ has led to almost total sky coverage over the primary and extended missions. However, this strategy limits the observational baseline, creating difficulties in detecting long-period planets. Simulation work by 41 ; 40 ; 39 ; 169 and most recently 143 has found that the majority of long-period planets detected via TESS will have unsolved periods, these being detected either from a single ‘monotransit’ or ‘duo/triotransits’ with two or three events separated by the \sim 2 yr gap between TESS visits ( 78 , e.g;) . In the former case the period is almost entirely unconstrained, although the duration of transit can give some indication, whilst in the latter case the true period forms part of a sequence of aliases, each an integer fraction of the gap between transits. This means extensive follow-up observational programs are required to confirm these planets and find their true orbital periods. The Next Generation Transit Survey ( 173 , NGTS;) is one of the best suited instruments for the photometric follow-up of unsolved TESS detections. It has already been used to confirm several long-period transiting giant planets ( 68 ; 65 ; 71 ; 14 ; 167 ; 166 ; 97 ) as well as brown dwarf and late M dwarf companions ( 64 ; 66 ; 67 ; 108 ; 81 ; 144 ) . The Characterising ExOPlanet Satellite ( 19 , CHEOPS ;) has similarly been used from space to confirm smaller transiting long-period planets ( 168 ; 128 ; 165 , e.g;) . Spectroscopic radial velocity campaigns using instruments such as CORALIE ( 137 ) , the Fiberfed Extended Range Optical Spectrograph ( 94 , FEROS;) or the High Accuracy Radial velocity Planet Searcher ( 134 ; 115 , HARPS;) are also essential in such follow-up efforts. These observations act either to confirm the mass of a planet with a photometrically determined period, or to provide an initial period that can be confirmed more precisely with additional photometry ( 152 ; 85 ; 53 ; 27 ; 20 , e.g.) . It is in this context that we present the discovery of TIC-65910228 b (which we also refer to as NGTS-38 b): a transiting giant planet on a 180.5 day orbit around a bright F type star. This makes it one of the longest period transiting planets ever found and one of the most amenable of these to characterisation. In Section 2 we describe the initial detection of the planet from a single TESS transit (Section 2.1 ), the detection of an egress with NGTS (Section 2.2 ), simultaneous observation of the most recent transit from several facilities (Section 2.3 ), and longer term photometric monitoring (Section 2.4 ) used to search for rotational signals. In Section 3 we describe the spectroscopic campaigns with CORALIE (Section 3.1 ) and HARPS (Section 3.2 ) that allowed a mass measurement of the planet via radial velocity and spectroscopic characterisation of the host star, which we describe in Section 4 . In Section 5 we describe the joint global model fit we performed on the previously described data in order to obtain a full orbital solution for NGTS-38 b. Then, in Section 6 , we present the parameters derived from our fit alongside some interior structure modelling and discuss their significance. We then discuss the potential for future work on the system in Section 7 before summarising our work with some concluding remarks in Section 8 . 2 Photometric Observations The photometric datasets for TIC-65910228  / NGTS-38 are summarised in Table 1 and they are described in the remainder of this section. Table 1 : Summary of photometric observations of NGTS-38. Facility Night(s) Observed (UTC) N_{\text{images}} * Exptime (s) N_{\text{nights}} ** Filter Pipeline N_{\text{telescopes}} Note(s) WASP-South 2011 Nov 2 - 2012 Mar 29 2464 30 77 WASP WASP 1 ASAS-SN 2016 Feb 2 - 2018 Sep 17 795 90 276 V ASAS-SN 1 ASAS-SN 2018 Aug 8 - 2025 Nov 23 4544 90 1304 g ASAS-SN 1 TESS -S07 2019 Jan 8 - 2019 Feb 16 1086 1800 22.6 TESS TESS -SPOC 1 TESS -S33 2020 Dec 18 - 2021 Jan 13 3485 600 24.2 TESS TESS -SPOC 1 a TESS -S34 2021 Jan 14 - 2021 Feb 8 3358 600 23.2 TESS TESS -SPOC 1 TESS -S61 2023 Jan 18 - 2023 Feb 12 9503 200 23.0 TESS TESS -SPOC 1 TESS -S87 2024 Dec 18 - 2025 Jan 14 14576 120 20.2 TESS SPOC 1 TESS -S88 2025 Jan 14 - 2025 Feb 11 19432 120 27.0 TESS SPOC 1 NGTS 2023 Aug 25 - 2024 Feb 4 118315 10 115 NGTS bsproc 1 NGTS 2024 Apr 16 - 2024 May 12 3012 10 8 NGTS bsproc 1 NGTS 2024 Aug 18 - 2024 Nov 21 44971 10 69 NGTS bsproc 1 NGTS 2024 Nov 22 - 2024 Dec 31 131761 10 36 NGTS bsproc 2 b NGTS 2025 Dec 2 - 2025 Dec 5 22903 10 4 NGTS bsproc 4 LCO-SSO 1m 2025 Dec 4 900 10 1 i ′ AstroImageJ 2 c LCO-SSO 0.35m 2025 Dec 4 1106 27 1 i ′ AstroImageJ 2 c LCO-CTIO 1m 2025 Dec 4 1068 10 1 i ′ AstroImageJ 2 d LCO-CTIO 0.35 2025 Dec 4 652 27 1 i ′ AstroImageJ 1 d El Sauce 2025 Dec 4 751 32 1 Rc AstroImageJ 1 d DSC0.4m 2025 Dec 4 120 135 1 Chroma R AstroImageJ 1 d OACC - CAO (0.6 m) 2025 Dec 4 196 100 1 i ′ 2 AstroImageJ 1 d • * For TESS N_{\text{images}} is counted as the number of individual points in the lightcurve with a data quality flag of 0. • ** For TESS this is calculated by (Exptime (s) \times N_{\text{images}})/86400 s while for all ground based instruments this is simply the count of nights with observations. • a: Full Transit on UTC 2020 December 25. b: Transit Egress on UTC 2024 December 8. c: Transit Ingress on UTC 2025 December 4. d: Transit Egress on UTC 2025 December 4 Figure 1 : TESS target pixel file for NGTS-38 in sector 33. Each red circle represents a Gaia source down to a magnitude difference of 6 from the target. The circles are also sized according to the difference in magnitude from the target, which is marked with a white x and numbered 1. Figure 2 : Discovery photometry for NGTS-38 b. All TESS lightcurves have been binned to 30 minute cadence for ease of comparison. NGTS data has been binned to nightly cadence and outliers with flux <0.998 or flux errors >0.001 removed. Red vertical dashes along the bottom x-axis show the expected transit times. Similar green dashes are shown on the inside of the top x axis to show transits actually covered by the data. The transit predicted at BJD-2457000 \approx 3292 does fall within the time limits of the first run of NGTS data but falls within a data gap. (a) TESS Sector 33 transit. (b) NGTS transit. Figure 3 : Discovery transit lightcurves of NGTS-38 b. Each panel shows the median transit model (see Section 5 ) as a solid red line. The transit data plotted is shown as black square markers with errorbars. For NGTS the data are binned to 10-minute cadence to match TESS and the unbinned data are shown as gray dots. The top panel of each subfigure shows the data and model while the lower panel of each shows the residuals after the model has been subtracted from the observed data. The TESS data shows a systematic ‘bump’ in flux post transit, this is probably an artefact of the PDCSAP_FLUX flattening of the lightcurve. Our own spline fit (see Section 5 ) failed to remove this but the impact on any fitted parameters is likely to be negligible. (a) LCOGT 1m (b) LCOGT 0.35m (c) DSC0.4m (d) OACC - CAO (0.6m) (e) El Sauce Figure 4 : Simultaneous transit lightcurves of NGTS-38 b from the night beginning 2025 December 4 UTC (T 10 =2461014.5077 BJD) Each panel shows the median transit model (see Section 5 ) as a solid red line. The top panel of each subfigure shows the data and model while the lower panel of each shows the residuals after the model has been subtracted from the observed data. Unbinned data is shown as gray dots while the black squares with error bars represent the data binned to 10 minute cadence. The LCOGT data for the same diameter telescopes from CTIO and SSO have been stitched together to save space but these data were fitted separately. 2.1 TESS TESS has been operating a (near) all-sky survey since 2018 aimed at discovering exoplanets transiting bright host stars. TESS ’ four \sim 10 cm cameras are arranged vertically to provide a rectangular 24\degree\times 96\degree field of view. TESS rotates this field of view around the sky in ‘sectors’, which are each observed for \sim 27 days at a time, although overlap between sectors can provide some targets with longer baselines. NGTS-38 was observed by TESS in sectors 7, 33, 34, 61, 87 and 88. We show the sector 33 Target Pixel File created using tpfplotter ( 7 ) in Figure 1 . In the most recent two sectors, lightcurves were produced onboard the spacecraft at two minute cadence by the Science Processing Operations Centre ( 92 , SPOC;) . The lightcurves for the remaining sectors were produced from the Full Frame Images (FFIs) using the TESS -SPOC pipeline ( 30 ) at cadences of 30-minutes for sector 7, 10-minutes for sectors 33 and 34 and 200-seconds for sector 61. We performed quality cuts on all of these lightcurves, only using data points with a quality flag of 0. We used the PDCSAP_FLUX column for all sectors. These lightcurves are plotted in Figure 2 . A single transit of NGTS-38 b was detected at BJD=2459209.23 (see Figures 2 and 3(a) ) in the Sector 33 TESS -SPOC FFI lightcurve by the Single Transit Search (STS) pipeline, a custom algorithm from the TESS Single Transit Planet Candidate Working Group (TSTPC WG). The TSTPC WG uses STS to systematically search TESS -SPOC lightcurves for single transit events, aiming to increase the yield of TESS long-period planets. The STS algorithm, employing astropy Box-Least Squares method ( 11 ) , identified the event by searching for periods near or beyond the TESS sector length across a thousand trial durations from 1 hour to 1 day. The transit signal was then compared against a sinusoidal model to exclude the possibility of stellar variability. The resulting data validation report was twice independently vetted, followed by a group evaluation session. The field is crowded (seen in Figure 1 ), with numerous faint sources within the photometric aperture. However, pixel-level lightcurves generated with eleanor ( 58 ) confirmed that the transit source is on-target on NGTS-38. The contaminating sources are too faint ( \Delta\text{m}_{{\it TESS}}>4 ) to produce the observed \sim 3 parts per thousand (ppt) transit depth, and no significant centroid motion was detected during the event, ruling them out as the event’s origin. The validated planet candidate parameters were uploaded to ExoFOP-TESS as the final step of TSTPC’s processes. The event in Sector 33 was also detected by the monofind algorithm ( 64 ; 78 ) , independently from and contemporaneously with the TSTPC WG detection. 2.2 NGTS NGTS is a ground based photometric facility made up of twelve individual robotic 20cm telescopes located in Paranal, Chile. Despite being ground based, the precision of NGTS is able to rival or exceed that of TESS ( 29 ; 129 ; 16 ) , making it very well suited to photometric follow-up of TESS single transit candidates ( 15 ) such as NGTS-38. NGTS-38 was observed by NGTS in four separate observing campaigns for a total of 228 nights between August of 2023 and January of 2025, these are summarised in Table 1 . The first campaign consisted of 115 nights of observation between 2023 August 25 and 2024 February 4. The second brief campaign consisted of 8 nights between 2024 April 16 and 2024 May 12, this sparse coverage was due to technical issues with the NGTS roof control system and cloudy weather. The third campaign began on 2024 August 18 with 69 nights before 2024 November 21. These first three campaigns all used a single telescope while the fourth and final campaign used two simultaneously for 36 nights of observations beginning immediately after the previous on 2024 November 22 and lasting until 2024 December 31. These data were reduced using the bsproc pipeline as described in 29 using fixed comparison stars and apertures across each run to ensure night to night photometric stability. A sample of the reduced lightcurve data can be found in Table 4 . Before any analysis was done, we median normalised the data across each campaign. We searched the NGTS monitoring data for a second planetary transit using the algorithm described by 64 , which uses the initial TESS detection as a template in a cross-correlation with the NGTS photometry. This revealed a candidate egress event on the night of 2024 December 8, which was confirmed with a visual inspection of the data (see Figure 3(b) ). No other transit events were identified in the 228 nights of NGTS photometry. With two transits of NGTS-38 b detected, the orbital period of the planet was constrained to a finite set of alias of the time separation of the two transits, which was 1444 d. The majority of these aliases were ruled out by either the TESS or NGTS photometry, including all alias periods shorter than 72 d. The egress event on 2024 December 8 showed a strong systematic increase in flux, comparable to the transit depth, at the beginning of the night, which was coincident with observations at high airmass ( \gtrsim 1.4 ). To correct for this, we fit a linear model of flux, f , against airmass, \sec z , to the residuals from the template match model against the data from the night of the transit (2024 December 8). The function took the form f=C_{1}\sec z+C_{0} with the fit resulting in coefficients of C_{1}=0.00244519 and C_{0}=-0.00383989 . We launched a short fifth NGTS campaign from 2025 December 2 to 2025 December 5 with four telescopes to target the 180.5 d alias, which was the best match to our radial velocity observations (see Section 3 ). However, due to high cloud cover over Paranal at the time we were unable to obtain any usable data. 2.3 TFOP SG1 Observations After identifying the most likely orbital period for NGTS-38 b of 180.5 d using TESS and NGTS photometry (Sects. 2.1 & 2.2 ) and radial velocity observations (Sect. 3 ), we sent out a general call to the TESS Follow-up Observing Program Sub Group 1 ( 38 , TFOP SG1;) 2 2 2 https://tess.mit.edu/followup to target the predicted transit ingress on the night beginning 2025 December 4 UTC and the egress on 2025 December 5 UTC. Observations were contributed by the Las Cumbres Observatory Global Telescope Network ( 28 , LCOGT; Section 2.3.1 ;) TESS Follow-up Key Project (PI: Shporer), as well as three private observatories: The Deep Sky Chile 0.4 m (DSC 0.4 m; Section 2.3.2 ), the Osservatorio Astronomico di Campo Catino (OACC; Section 2.3.3 ) and the El Sauce Observatory (Section 2.3.4 . Photometric extraction was carried out for all of these instruments with the AstroImageJ package ( 37 ) . The comparison star ensemble for each lightcurve was selected by minimizing the RMS of the residuals of a simultaneous fit to a transit model and a linear fit to airmass using AstroImageJ . The resulting transit light curves are all shown in Figure 4 . The detection of this transit ruled out the majority of the remaining alias periods allowed by the TESS and NGTS transits (Sect. 2.2 and Fig. 3 ). The full set of aliases allowed by the combined photometry are: 72, 90, 120, 181 and 361 d. Only the 180.5 d alias is also consistent with our radial velocity observations (see Section 3 ). 2.3.1 LCOGT A transit of NGTS-38 b beginning on 2025 December 4 UTC was observed from two Las Cumbres Observatory Global Telescope network nodes. The transit ingress was observed simultaneously with two 1.0 m and two 0.35 m telescopes at the Siding Spring Observatory (SSO) node near Coonabarabran, Australia and the egress was observed simultaneously with two 1.0 m and one 0.35 m telescopes at the Cerro Tololo Inter-American Observatory (CTIO) node in Chile. The 1.0 m telescopes are equipped with 4096 \times 4096 SINISTRO Cameras, with an image scale of 0.389″ per pixel and a field of view (FOV) of 26^{\prime}\times 26^{\prime} . The 0.35 m Planewave Delta Rho 350 telescopes are equipped with 9576\times 6388 QHY600 CMOS cameras having an image scale of 0\aas@@fstack{\prime\prime}73 per pixel, resulting in a 114′\times 72′ full field of view. We used the optional 30′\times 30′ sub field of view for a faster detector read-out. All observations were carried out in the Sloan-i ′ filter with exposure times of 10 s and 27 s for the 1.0 m and 0.35 m telescopes, respectively. The image data were calibrated with the standard LCOGT banzai pipeline ( 117 ) , and photometric extraction was performed with AstroImageJ using uncontaminated photometric apertures of 6–9″. We combined all 1.0 m datasets in Figure 4(a) and all 0.35 m datasets in Figure 4(b) . 2.3.2 DSC 0.4m The 2025 December 5 UTC egress was also observed from near Ovalle, Chile, with the Deep Sky Chile 0.4 m (DSC 0.4 m) Planewave CDK400 telescope, which was equipped with a QHYCCD600M camera and Chroma R filter (650 nm \pm 50 nm). The 9576 x 6388 pixel detector has an image scale of 0.4" per pixel and a FOV of 62^{\prime}\times 42^{\prime} . The images were calibrated and photometric data were extracted with AstroImageJ using circular 5" photometric apertures. The egress lightcurve is shown in Figure 4(c) . 2.3.3 OACC - CAO 0.6m The 2025 December 5 UTC egress was also observed from the Osservatorio Astronomico di Campo Catino (OACC) Campocatino Austral Observatory (CAO) 0.6 m telescope (El Sauce, Atacama desert, Chile) equipped with a Moravian C3-61000EC PRO camera that provides a FOV of 30^{\prime}\times 20^{\prime} with a pixel scale of 0.2" per pixel. The images were aqcuired with a standard photometric Astrodon i ′ 2 filter. The time series, composed of 196 x 100s images, was calibrated and photometric data were extracted with AstroImageJ using circular 6.8" photometric apertures. The egress lightcurve is shown in Figure 4(d) . 2.3.4 El Sauce The transit egress was also observed from El Sauce observatory in Coquimbo Province, Chile on UTC 2025/12/05 with a CDK500 0.5 m robotic telescope equipped with an Rc filter and a Moravian C3-26000 camera with 6252 x 4176 pixels binned 2 x 2 in-camera resulting in an image scale of 0.449"/pixel. The 751 x 32 s images were calibrated and photometric data were extracted with AstroImageJ using circular 5.8" photometric apertures. The egress lightcurve is shown in Figure 4(e) . 2.4 Archival Photometry 2.4.1 WASP The Wide Angle Search for Planets ( 136 , WASP;) was a ground based transit survey with a northern facility (often simply referred to as WASP) at the Observatorio del Roque de los Muchachos on the island of La Palma in the Canary Islands and a Southern facility (WASP-South) at the Sutherland Station of the South African Astronomical Observatory. Both WASP instruments consisted of eight Canon Telephoto lenses with a \sim 11 cm aperture with custom 2048\times 2048 pixel CCD cameras on a fixed robotic mount, resulting in a wide 482 square degree field of view. NGTS-38 was observed by WASP-South from 2011 November 2 to 2012 March 29. The dataset contains no transit coverage, and the transit is shallower than WASP’s \sim 1\% detection limit. However, the WASP lightcurve was still used to attempt to measure a photometric rotation period of the star which we present in Section 4.4.1 . 2.4.2 ASAS-SN The All Sky Automated Survey for SuperNovae ( 103 , ASAS-SN) is made up of several stations hosted at various locations around the world by the LCOGT network (see Section 2.3 and 2.3.1 ). Each station consists of a single Nikon telephoto lens with a 14 cm aperture and an approximately 4.5 square degree field of view. ASAS-SN is a transient survey and so is not well suited for transit detection given the low number of observations per target on a given night, however, its long baseline of nightly coverage does allow for monitoring rotation and/or activity in stars ( 55 , e.g;) . NGTS-38 was observed by ASAS-SN from 2016 February 2 to 2018 September 17 in the V band and 2018 August 8 to 2025 November 23 in the g band. We apply the same quality cuts as in 55 to our dataset in order to use the ASAS-SN lightcurve to search for a stellar rotation signal, which we present in Section 4.4.1 . 3 Spectroscopic Observations (a) Radial velocities as a function of time. (b) Radial velocities after phase-folding on the orbital period. Figure 5 : Radial velocity measurements of NGTS-38 with the fitted radial velocity baseline offset values subtracted (see Section 5 and Table 7 ). Data points from CORALIE are denoted with black circles with those from the first epoch unfilled and those from the second epoch filled. HARPS data points are shown as blue squares. The median fitted RV model is also overplotted with a red solid while the best fitting model to the 361.0 day period alias is shown with an orange dashed line in the top panel. Subfigure (a) shows the data plotted versus time while (b) shows the same data as a function of orbital phase. For each subfigure, the top panel shows the data and model, while the bottom panel shows the residuals with the model subtracted. Figure 6 : Spectroscopic Lomb-Scargle periodograms of the radial velocity timeseries ( top ), the combined HARPS and CORALIE window function ( middle ) and \log\,\mathrm{R}^{\prime}_{\mathrm{HK}} measurements from HARPS ( bottom ). The window function was created by running a Lomb-Scargle periodogram on the RV timeseries with a constant y value of 1. False alarm probabilities of 1%, 0.1% and 0.01% are represented by horizontal black lines with dotted, dashed and dash-dotted styles, respectively. A solid orange vertical line represents the orbital period of the planet at \sim 180.5 d while dashed vertical orange lines represent the photometric period aliases at 361.0, 120.3, 90.3 and 72.2 days. Additionally purple dotted vertical lines are shown to represent periods of 1 year and half a year. In this section we describe the spectroscopic observations taken of TIC-65910228 / NGTS-38 with CORALIE ( 137 , Section 3.1 ) and the High Accuracy Radial velocity Planet Searcher (HARPS; 134 ; 115 , Section 3.2 ) . These spectra were used to produce time-series radial velocity measurements shown in Tables 5 and 6 , respectively, and are presented together in Figure 5 . Before any models were fitted to the data, we performed an initial search for periodic signals by creating a generalised Lomb-Scargle periodogram ( 112 ; 150 ; 179 ) of the combined CORALIE and HARPS radial velocity data using astropy ( 11 ) . To combine these datasets we subtracted the median from each set of RVs before concatenating them together. The resulting periodogram is shown in Figure 6 . The periodogram shows two peaks at \sim 180 and \sim 360 days which effectively rules out all but the 361 day and 180.5 day photometric period aliases (see Section 2 ). The relative strength of the 180.5 day peak compared to to the 361 day signal as well as visually inspecting and comparing the best fits at both periods (see Figure 5(a) and Section 5 ) led us to conclude that the 180.5 day alias was in fact the true period of the system. 3.1 CORALIE CORALIE is a fiber-fed echelle spectrograph with a resolution of \text{R}\sim 60\,000 , mounted on the 1.2-m Leonhard Euler Telescope at ESO La Silla Observatory in Chile. CORALIE is capable of achieving radial velocity precision on the order of \sim 10 m s -1 ( 137 ) , which makes it well suited to the detection of giant planets such as NGTS-38 b. On the night of 2023 April 16 a single spectrum of NGTS-38 was taken followed by an additional twelve points between 2023 December 10 and 2024 May 5, with a further 31 observations performed between 2024 October 21 and 2025 June 8 with 44 total spectra taken. All the observations of NGTS-38 taken with CORALIE had an exposure time of 1200-s and were taken at airmasses between 1.00 and 1.74 with the resulting spectra having signal-to-noise ratios (SNRs) between 19.6 and 40.4 in spectral order 62. The spectra were reduced using the standard CORALIE Data Reduction System (DRS) 3.8 with a G2 mask used for cross correlation that produced the radial velocity measurements shown in Table 5 and Figure 5 . Due to an instrument update for CORALIE between the 2nd and 3rd set of observations described above, we treat the data from before and after this point as two separate instruments (CORALIE 14 and CORALIE 24 , respectively) for the purpose of fitting instrumental offsets and jitter terms (see Section 5 ). 3.2 HARPS Alongside the observations made with CORALIE, we also observed NGTS-38 with HARPS; a fiber-fed echelle spectrograph with a spectral resolution of \text{R}\sim 115000 , which is mounted on the 3.6-m telescope at the ESO La Silla Observatory. The greater telescope diameter and spectral resolution of HARPS allow for more precise radial velocity measurements than CORALIE, with HARPS capable of reaching \sim 1 m s -1 precision. The HARPS observations of NGTS-38 were taken in a very similar time-frame to the CORALIE data. Nine spectra were taken between 2023 December 20 and 2024 March 11, with a further 13 between 2024 October 14 and 2025 May 6. An additional HARPS spectrum, taken on 2026 January 5, was incorporated into the RV dataset but not used in the determination of the host star’s spectral parameters (see Section 4.2 ). The first seven of these observations used an exposure time of 1500-s and were taken at airmasses between 1.02 and 1.59 with a resulting SNR of 51.4-78.4 in order 64. The remaining spectra were all taken with an exposure time of 1200-s at airmasses between 1.00 and 1.49 with resulting SNRs between 35.3 and 56.1 in order 64. These data were reduced with DRS 3.3.6 using an F9 mask. The resultant radial velocity values as well as activity indicators, including the S index derived \log\,\mathrm{R}^{\prime}_{\mathrm{HK}} value are shown in Table 6 and Figure 5 . 4 Host star analysis 4.1 Catalogue parameters Catalogue parameters for TIC-65910228 / NGTS-38 are shown in Table 2 . These were taken from the 2 Micron All Sky Survey ( 157 , 2MASS;) , Gaia data release 3 ( 63 , Gaia DR3;) , the TESS Input Catalogue v8 ( 159 , TIC 8;) , the AAVSO Photometric All Sky Survey ( 80 , APASS;) , the Wide field Infrared Survey Explorer ( 177 , WISE;) , Tycho ( 87 ) and SkyMapper ( 95 ) Data Release 4 ( 127 , DR4;) catalogues where specified. We use the gaiadr3-zeropoint package to calculate a zero point offset to the parallax as described in 111 ; 110 , finding a value of of -0.01592 mas which we apply to the Gaia DR3 parallax of 3.7400\pm 0.0128 mas resulting in the final value we quote in Table 2 . To aid in visualising the colour and magnitude information from these catalogues we overplot NGTS-38 on a Gaia DR3 Hertzsprung-Russell diagram ( 82 ; 147 , HRD;) of all TESS -SPOC target stars from 52 in Figure 7 . The star’s position on the HRD shows a colour and luminosity broadly consistent with a main sequence F6-F7 type star, albeit a mildly over-luminous one. Figure 7 : Gaia DR3 Hertzsprung-Russel diagram for all TESS -SPOC target stars from 52 . NGTS-38 is overplotted as a white star with black borders. A solar metallicity 2.2 Gyr MIST isochrone ( 51 ; 34 ) is overplotted as a solid blue line. The star sits significantly above the isochrone due to its enhanced metallicity giving it a larger radius (see Sections 4.3 and 6.1 ). Table 2 : NGTS-38 stellar parameters. Property Value Source Identifiers Tycho ID 6536-3043-1 Tycho(§ 4.1 ) SkyMapper DR4 ID 35585731 SkyMapper (§ 4.1 ) 2MASS ID J07145152-2925498 2MASS (§ 4.1 ) Gaia Source ID 5606317297918628992 Gaia DR3 (§ 4.1 ) TIC ID 65910228 TIC 8 (§ 4.1 ) Coordinates RA (hh:mm:ss.ss) 07:14:51.52 Gaia DR3 (§ 4.1 ) DEC (dd:mm:ss.ss) -29:25:49.91 Gaia DR3 (§ 4.1 ) Proper motion and parallax \mu_{\text{RA}} (mas y -1 ) 13.5664\pm 0.0108 Gaia DR3 (§ 4.1 ) \mu_{\text{DEC}} (mas y -1 ) -21.0834\pm 0.0131 Gaia DR3 (§ 4.1 ) Parallax (mas) 3.75592\pm 0.01280 Gaia DR3 (§ 4.1 ) Systemic velocity (km s -1 ) 19.9306\pm 0.0014 This work (§ 5 ) Galactic Kinematics U (km s -1 ) 18.407\pm 0.096 This work (§ 4.5 ) V (km s -1 ) -32.8103\pm 0.054 This work (§ 4.5 ) W (km s -1 ) 0.544\pm 0.019 This work (§ 4.5 ) v_{\text{tot}} (km s -1 ) 37.62\pm 0.07 This work (§ 4.5 ) e_{\text{gal}} 0.16 This work (§ 4.5 ) J_{z} (kpc km s -1 ) 1608.6 This work (§ 4.5 ) Magnitudes V (mag) 10.230\pm 0.020 APASS (§ 4.1 ) B (mag) 10.738\pm 0.026 APASS (§ 4.1 ) u (mag) 11.703\pm 0.011 SkyMapper (§ 4.1 ) v (mag) 11.300\pm 0.011 SkyMapper (§ 4.1 ) g (mag) 10.334\pm 0.009 SkyMapper (§ 4.1 ) r (mag) 10.116\pm 0.010 SkyMapper (§ 4.1 ) i (mag) 10.047\pm 0.017 SkyMapper (§ 4.1 ) z (mag) 10.051\pm 0.012 SkyMapper (§ 4.1 ) G (mag) 10.1103\pm 0.0002 Gaia DR3 (§ 4.1 ) BP (mag) 10.3797\pm 0.0003 Gaia DR3 (§ 4.1 ) RP (mag) 9.6821\pm 0.0002 Gaia DR3 (§ 4.1 ) TESS (mag) 9.734\pm 0.006 TIC 8 (§ 4.1 ) J (mag) 9.251\pm 0.026 2MASS (§ 4.1 ) H (mag) 9.010\pm 0.026 2MASS (§ 4.1 ) K (mag) 8.938\pm 0.023 2MASS (§ 4.1 ) W1 (mag) 8.909\pm 0.023 WISE (§ 4.1 ) W2 (mag) 8.947\pm 0.021 WISE (§ 4.1 ) W3 (mag) 8.912\pm 0.027 WISE (§ 4.1 ) Spectral Parameters v_{\rm mic} (km s -1 ) 1.4\pm 0.2 This Work (§ 4.2 ) v_{\rm mac} (km s -1 ) 7.6\pm 0.8 This Work (§ 4.2 ) v\sin i_{\star} (km s -1 ) 12\pm 3 This Work (§ 4.2 ) T_{\text{eff}} (K) 6310\pm 129 This Work (§ 4.2 ) [Fe/H] (‘dex’) 0.33\pm 0.09 This Work (§ 4.2 ) \log g ( \log(\text{cgs}) ) 4.06\pm 0.02 This Work (§ 4.3 ) \log\,\mathrm{R}^{\prime}_{\mathrm{HK}} -5.07\pm 0.07 This Work (§ 4.4.2 ) Derived parameters \rho_{\star} (g cm -3 ) 0.302_{-0.030}^{+0.035} This Work (§ 5 ) R ⋆ ( R ⊙ ) 1.88\pm 0.08 This Work (§ 4.3 ) M ⋆ ( M ⊙ ) 1.46\pm 0.09 This Work (§ 4.3 ) Age (Gyr) 2.2\pm 0.5 This Work (§ 4.3 ) Distance (pc) 265\pm 1 This Work (§ 4.3 ) 4.2 Spectral fit To fit spectral parameters for the host star we co-added the first 22 HARPS and 44 CORALIE spectra into two combined spectra with SNR values of 151 and 112, respectively. We then analysed these spectra using Parameters Approximated With Synthesis ( 61 ; 62 , paws ;) . paws uses the functions provided in the iSpec package ( 21 ; 22 ) to determine the atmospheric parameters via the equivalent widths (EWs) of the Fe I and Fe II lines, assuming ionization and excitation equilibrium. The resulting values from this method were then used as priors for a spectral synthesis fit to obtain the final stellar parameters. We calculated an inverse variance weighted mean of the results from both the CORALIE and HARPS spectra, obtaining values of T_{\text{eff}} , [Fe/H], v_{\rm mic} , v_{\rm mac} , and v\sin i_{\star} that we adopt in Table 2 . Note that we do not adopt the spectroscopically derived \log g of 3.8\pm 0.1 from paws as such values from spectra are known to be unreliable due to strong correlations with T_{\text{eff}} and [Fe/H] ( 163 ; 121 ) . 4.3 Isochrone fit To determine the final set of stellar parameters we adopt in Table 2 , we fitted the stellar models from the MESA ( 131 , Modules for Experiments in Stellar Astrophysics;) Isochrones and Stellar Tracks ( 51 ; 34 , MIST;) using the isochrones package ( 122 ) . In order to derive radius, mass (and thus isochronal \log g ), and age, we used the optical B and V band magnitudes as well as the J,H,K,W1,W2 and W3 magnitudes to cover a wider range of the spectral energy distribution. All of these magnitudes are listed in Table 2 . We furthermore used the values of T_{\text{eff}} and [Fe/H] from paws (see Section 4.2 ) and the zero point corrected Gaia DR3 parallax applied listed in Table 2 . Finally, we also used the host density ( 0.302_{-0.030}^{+0.035} g cm -3 ) derived from preliminary fits to the transit data with no stellar parameters inputted (see Section 5 ). We note that the final result did not depend, within errors, on the inclusion or exclusion of the stellar density derived from the transit. Final values and precision uncertainties were extracted as the median value and the 16th and 84th percentiles of the posterior distributions. We adopt the recommendations of 161 to account for the inherent uncertainties between stellar models; adding an additional 0.042% error term in quadrature to the host radius. In addition we made use of the kiauhoku package ( 36 ) to compare fits of the spectroscopic T_{\text{eff}} , \log g and [Fe/H] against MIST, Dartmouth ( 50 ) , Yale Rotating stellar Evolution Code ( 47 , YREC;) and Garching STellar Evolution Code ( 172 , GARSTEC;) stellar models. We calculated the standard deviation across the values fitted against each model and added these in quadrature to our values. This error term for \log g was so small as to be negligible (\sim 10^{-7}) . However, we used terms for the stellar mass and age of 0.08 M ⊙ and 0.4 Gyr, respectively. With these error terms added this resulted in the final stellar parameters we adopt in Table 2 . 4.4 Activity and rotation 4.4.1 Rotation Using eequation 1 from 171 and our value of v\sin i_{\star} from paws (see Section 4.2 ), we estimate an upper limit on the stellar rotation period of 8\pm 2 d. We attempt to measure a photometric rotation period by creating Lomb-Scargle periodograms on the largest consecutive runs of each of our datasets. For NGTS we use the first run between 2023 August 25 - 2024 February 4 and for TESS we separately median normalised and then stitched together the SPOC lightcurves (using the SAP_FLUX column) for sectors 87 and 88. We also used the entire WASP and ASAS-SN datasets (See Section 2.4.1 and 2.4.2 ), separating the latter by V and g bands. The resulting periodograms are shown in Figure 8 . Figure 8 : Lomb Scargle periodograms of TESS sectors 87 and 88 ( top ), the first run of NGTS data ( upper middle ), ASAS-SN in both the V band ( middle ) and g band ( lower middle ) and WASP ( bottom ) for NGTS-38. False alarm probabilities of 1%, 0.1% and 0.01% are represented by horizontal black lines with dotted, dashed and dash-dotted styles respectively. For TESS , its 13.7 d orbital period is marked with a vertical red dashed line as well as aliases at 2\times (a TESS sector duration) and 0.5\times this period (which coincides with the mid-orbit gap in newer sectors). The TESS periodogram shows two peaks of interest, one at 6.7 d and another broad peak around \sim 20-30 d. However, these are close to the mid orbit gap and mid sector gap, respectively, and should be interpreted cautiously. The NGTS periodogram is nosier than TESS but still shows a few peaks of interest. The peak at \sim 1 d is almost certainly due to the nightly cadence of the data. There is a peak at 6.9 d similar to that found in TESS , with another at 26.9 d which coincides with the lunar period of 27.3 d and thus is likely caused by dilution from moonlight modulating with the lunar phases. Another peak appears at 93 d, however given this is approaching the baseline it should also be interpreted cautiously. A peak at \sim 7 d appears in both TESS and NGTS photometry and is in agreement with the v\sin i_{\star} derived rotation period – lending some credence that this may be the true rotation period of the star. However, phase folding on this period doesn’t show a clear sinusoidal signal for either dataset. Additionally, we do not see this peak in either ASAS-SN bandpass, seeing only a \sim 300 d peak in both likely due to the Earth’s orbit and another at 29.5 d in the g band likely due to the moon. The WASP periodogram also shows no convincing peaks at any period beyond \sim 1 d. Hence we conclude that, in spite of a tentative \sim 7 d signal, we do not recover any significant rotation period for NGTS-38 photometrically. 4.4.2 \log\,\mathrm{R}^{\prime}_{\mathrm{HK}} We calculated an inverse variance weighted mean and standard deviation of all the individual values of \log\,\mathrm{R}^{\prime}_{\mathrm{HK}} calculated by the DRS from each HARPS spectrum (see Table 6 ), finding of \log\,\mathrm{R}^{\prime}_{\mathrm{HK}} = -5.07\pm 0.07 . This indicates a low level of stellar activity, albeit not unusually so for an F6-F7 star ( 90 ; 89 ; 91 , See \log\,\mathrm{R}^{\prime}_{\mathrm{HK}} catalogues;) . A potential long-term downward trend in \log\,\mathrm{R}^{\prime}_{\mathrm{HK}} is visible. This trend is, however, unlikely to be genuine as the lower \log\,\mathrm{R}^{\prime}_{\mathrm{HK}} data points have systematically poorer precision and come from lower signal to noise spectra. We still attempted to investigate this by producing a periodogram of the \log\,\mathrm{R}^{\prime}_{\mathrm{HK}} data, which we present in Figure 6 . The periodogram shows a potential long term trend with a high plateau in power beyond 10^{3} d, although this could be driven by the lower SNR spectra taken later in the campaign. Also notable are the peaks that appear to coincide with those in the RV periodogram at \sim 180 and \sim 360 . There is also a weak negative correlation between the RVs and \log\,\mathrm{R}^{\prime}_{\mathrm{HK}} . Although, it is highly unlikely that the RV signal is activity induced due to its high amplitude. These peaks are instead likely an artefact of the spectroscopic window function. This is further supported by the fact that we see no similar peaks or RV correlations in other activity indicators such as the Full Width Half Maximum (FWHM) and Bisector Inverse Slope (BIS). 4.5 Galactic Kinematics and population membership Using the proper motions and parallax reported in Gaia DR3 and the HARPS systemic RV (see Sections 3 , 5 and Table 2 ), we compute the kinematic properties of NGTS-38 following 93 ; 17 and report U, V and W in Table 2 . As laid out in 17 ; 18 ; 140 , we compute the kinematic membership probability by correcting for the Local Standards of Rest reported in 105 ; 153 ; 42 ; 24 ; 60 ; 162 and 8 , also using the velocity dispersions and stellar fractions reported in 18 ; 33 . This results in a weighted kinematic thin disc probability of 98.2% for NGTS-38. This is consistent with the star’s enriched metallicity as thin disc stars are typically more metal rich than those in the thick disc ( 17 ; 18 ) . 5 Orbital solution We performed a joint model fit to the combined photometric and spectroscopic radial velocity dataset using the allesfitter package ( 75 ; 76 ) , which packages together various fitting code suites with the ellc modelling package ( 114 ) . We opted to use nested sampling ( 155 ; 156 ) via the dynesty package ( 158 ) contained within allesfitter . allesfitter uses 8 main fitted astrophysical parameters to describe the orbital solution of a system. The first and simplest of these is the orbital period, P , in days. Transit timings are described largely by this as well as the transit epoch, T_{0} . The transit shape comes from the radius ratio between the planet and host star, R p / R ⋆ , the combined host and planet radius relative to the orbital semi-major axis, (\mbox{R${}_{\star}$}+\mbox{R\textsubscript{p}})/a , and the cosine of the orbital inclination, \cos i . Radial velocity data solely determines the semi-amplitude, K , in km s -1 and also largely controls the orbital eccentricity ( e ) and angular argument of periastron ( \omega ), parametrized as \sqrt{e}\sin\omega and \sqrt{e}\cos\omega . On top of these astrophysical parameters, a wide array of instrumental parameters were also fitted to the data. For photometry we included logarithmic error terms, \log\sigma per instrument, which we initially set with wide uniform priors between -14 and 0. For all ground based photometry instruments we fit a simple flux baseline offset term with a uniform prior between \pm 0.001 . For TESS we performed a least squares minimisation spline fit to the baseline at each sampling step using allesfitter ’s built in hyrbrid_spline method. We modelled limb darkening using a quadratic law, for which allesfitter uses the q_{1},q_{2} parametrization from 101 , varying the priors between 0 and 1 for all instruments in all fits with the exception of the final fit described later in this section. The limb-darkening parameters were coupled together for the four LCOGT instruments. For the radial velocity data, a logarithmic jitter term, \log\sigma , was fitted per instrument with initial uniform priors, as well as an RV baseline offset. The first fit we performed was only to the TESS and radial velocity data. We allowed the period to vary uniformly between 179 and 181 days, based on the most likely peak in the RV periodogram (see Figure 6 ). Similarly, the transit epoch had a uniform prior distribution of a 2 day range centred on the reported TESS epoch on BJD=2459209.228. Wide uniform priors of \mathcal{U}(0,0.1) , \mathcal{U}(0,0.05) were placed on R p / R ⋆ and (\mbox{R${}_{\star}$}+\mbox{R\textsubscript{p}})/a , respectively, based on visual inspection of the TESS transit. The radial velocity semi-amplitude was allowed to vary between 0 and 500 m s -1 . \cos i was allowed to vary across its full possible range between 0 and 1 and the parameters of \sqrt{e}\sin\omega and \sqrt{e}\cos\omega were similarly sampled across their full possible ranges between -1 and 1. The RV baseline priors were set between 19.95 and 20.05 km s -1 for both CORALIE epochs and between 19.9 and 20 km s -1 for HARPS, in both cases based on a visual inspection of the data. All RV instrumental jitter terms had uniform priors between -7.6 and -4.6 \ln km s -1 . The transit parameters resulting from this fit were used for the template matching used to detect the NGTS egress described in Section 2.2 . After the detection of the NGTS egress (see Section 2.2 ), the detrended NGTS lightcurve was incorporated into a fit. The period of 180.531\pm 0.001 d and epoch of \text{BJD}=2459209.2277\pm 0.0032 from this fit were used to schedule the simultaneous set of additional transit observations described in Section 2.3 . We performed one final preliminary on the data incorporated the entire dataset and was performed to determine the transit derived stellar density used in the host star characterisation (see Section 4.3 ). As with all the previous fits, no stellar parameters were inputted to allesfitter to avoid any biasing of the resulting stellar density value. In addition we conducted a second fit with the same setup except the period prior distribution was uniform between 360 and 362 days instead of 179 and 181 days. Whilst the \mathcal{U}(179,181) day period prior fit provided a good fit to the data, the \mathcal{U}(360,361) day fit was much poorer (see Figure 5(a) ), effectively allowing us to rule out this period. The final fit was largely identical to the penultimate fit except for the addition of the stellar radius, mass and effective temperature into the fit, which allowed the determination of derived parameters such as the planet mass and radius. We also allowed allesfitter to use these stellar parameters to set an external prior on the stellar density. Solutions where the derived host density did not match this prior were penalised. In addition, we used the Limb Darkening Tool Kit ( 130 , ldtk ;) package to determine priors on the limb darkening coefficients. Since, ldtk returns limb darkening coefficients in the usual u_{1},u_{2} parametrization, we converted these values back to the 101 q_{1},q_{2} parameters in order to be usable in allesfitter . ldtk uses values of the host star T_{\text{eff}} , [Fe/H] and \log g along with filter response functions for each instrument. For TESS , NGTS and LCOGT exact filter response functions were available while for the El Sauce, DSC0.4m and OACC-CAO instruments we used generic Cousins R, Johnson R, and Sloan i ′ filters respectively from the Spanish Virtual Observatory filter profile service, 3 3 3 Available at https://svo2.cab.inta-csic.es/theory/fps/ assuming these to be close enough approximations. In these cases we widened the spread of the normal priors on both limb darkening coefficients by a factor of 3 to account for the approximation. The prior and posterior distributions of all parameters in this fit are shown in Table 7 . Also presented are the derived parameters using the inputted stellar parameters, shown in Table 8 . 6 Results 6.1 The TIC-65910228 / NGTS-38 system Table 3 : NGTS-38 b properties Parameter Value Source Period (days) 180.52791\pm 0.00038 allesfitter Transit Parameters T 0 (BJD) 2459209.2286\pm 0.0031 allesfitter R p / R ⋆ 0.05909\pm 0.00061 allesfitter R ⋆ /a 0.01258\pm 0.00035 allesfitter b_{\text{transit}} 0.8540_{-0.010}^{+0.0090} allesfitter b_{\text{occulation}} 0.59\pm 0.03 This work (§ 7.2 ) T 14 (hrs) 14.88\pm 0.13 allesfitter T 23 (hrs) 9.38_{-0.29}^{+0.27} allesfitter RV parameters K (km s -1 ) 0.1372\pm 0.0018 allesfitter e 0.308\pm 0.011 allesfitter \omega (degrees) 278.11\pm 0.70 allesfitter Derived parameters R p ( R J ) 1.081\pm 0.047 allesfitter M p ( M J ) 4.78_{-0.37}^{+0.40} allesfitter \rho_{\text{p}} (g cm -3 ) 4.69_{-0.78}^{+0.95} allesfitter \log g ( \log(\text{cgs}) ) 3.98\pm 0.03 allesfitter a (au) 0.695\pm 0.036 allesfitter Periapsis (au) 0.481\pm 0.026 a(1-e) Apoapsis (au) 0.91\pm 0.05 a(1+e) T_{\text{eq}} (K) 458\pm 11 allesfitter T_{\text{eq}} ;Apoapsis (K) 400\pm 10 This work (§ 6.1 ) T_{\text{eq}} ;Periapsis (K) 550\pm 14 This work (§ 6.1 ) R_{\text{Hill}} ( R J ) 147\pm 9 This work (§ 7.4 ) d_{\text{Roche}} ( R J ) 3.50\pm 0.28 This work (§ 7.4 ) TIC-65910228 / NGTS-38 is a bright (Vmag= 10.230\pm 0.020 ) star with a temperature of 6310\pm 129 K, which is consistent with an F6V-F7V type star ( 132 ) . While the star may superficially appear evolved due to its large radius ( 1.88\pm 0.08 R ⊙ ) compared to its mass ( 1.46\pm 0.09 M ⊙ ) and resulting low surface gravity ( \log g = 4.06\pm 0.02 ), the star is still on the main sequence. In evolutionary models of F type stars ( 51 ; 34 , e.g.) , the stellar radius is fairly dependent on metallicity as a higher metallicity leads to higher opacities in the upper stellar atmosphere. Hence, a metal rich star like NGTS-38 ([Fe/H]= 0.33\pm 0.09 ) can have an apparently inflated radius before leaving the main sequence (See the isochrone in Figure 7 . This is reflected in the isochrone derived stellar age of 2.2\pm 0.5 Gyr we find for NGTS-38, which is expected of an F6-F7 type star on the main sequence. The system is host to a transiting planet companion, NGTS-38 b, for which we list some properties in Table 3 . We find the planet to have a radius similar to that of Jupiter ( R p = 1.081\pm 0.047 R J ) but almost five times the mass ( M p = 4.78_{-0.37}^{+0.40} M J ), placing it firmly among the rare ‘super-Jupiter’ class of exoplanets. The planet is on a long-period (P= 180.52791\pm 0.00038 d), wide separation ( a= 0.695\pm 0.036 au) orbit with a moderate eccentricity ( e= 0.308\pm 0.011 ). Its wide orbital separation gives NGTS-38 b a relatively cool equilibrium temperature for a transiting exoplanet with T_{\text{eq}} = 458\pm 11 K at its semi-major axis. It should be noted, however, that a planet with as long a period and as eccentric an orbit as NGTS-38 b will experience temperature fluctuations throughout its orbit. Using the same assumptions used by allesfitter to estimate T_{\text{eq}} at a (those being a Jupiter like albedo of 0.3 and emissivity of 1) we calculate the equilibrium temperature at periastron and apastron, which we list in Table 3 . 6.1.1 Tidal interactions For a planet as distant from its host as NGTS-38 b, tidal interactions are expected to be negligible. To test this, we first calculate the tidal circularisation timescale using equation 4 from 88 . We use a range of values for the planetary tidal quality factor 10^{4}\leq Q_{\text{P}}\leq 10^{7} . All the resulting timescales are between \sim 10^{5}-10^{8} Gyr, more than the current age of the universe and far longer than both the current age and expected lifespan of the host star. As such we conclude that the planet’s eccentricity has not been significantly damped by tidal interactions with the host star in its lifetime and the planet will not experience any significant damping of its eccentricity until the host star begins post main sequence evolution. An accurate tidal locking timescale for a non-rigid body such as NGTS-38 b is a more complex affair ( 48 , see;) and beyond the scope of this paper. However, we can obtain a reasonable estimate using equation 1 from ( 74 ) assuming an initial rotation rate equal to that of Jupiter (\omega_{\text{P}}=1.7\times 10^{-4}\,\text{s}^{-1}) and varying Q_{\text{P}} again between 10^{4} and 10^{7} . The resulting values are on the order of 10^{60} Gyr, essentially meaning the planet will never be tidally locked during the host star’s main sequence lifetime. 6.2 NGTS-38 b in a population context NGTS-38 b is one of the longest period transiting planets ever discovered being in the 97.5th percentile of all transiting planets in terms of orbital period ( 2 ; 35 , NASA exoplanet archive;) . Within this 97.5th period percentile it is also one of the most amenable to characterisation with its host star belonging to the 90th percentile in V band magnitude. NGTS-38 b is also one of only thirteen planets with orbital periods >100 d to be discovered by the TESS mission at the time of writing, being the 7th longest period of these. The mass of NGTS-38 b is just within the sparsely populated mass regime between \sim 4 M J and \sim 10 M J , proposed to represent the transition from core accretion to gravitational instability as the primary mechanism of planet formation ( 148 ; 151 ; 124 ) . NGTS-38 b’s presence in this mass regime may be more intriguing given the enriched metallicity of its host star since 124 find the occurrence rate of planets with \mbox{M\textsubscript{p}}\geq 4\mbox{M\textsubscript{J}} decreases with increasing host metallicity (although the occurrence of planets with periods longer than 10 days increases). Additionally the moderate eccentricity of NGTS-38 b ( e = 0.308\pm 0.011 ) is consistent with the peak in the eccentricity of super-Jupiters at e\sim 0.3 identified by 23 . 6.2.1 Archive sample We retrieved a sample of transiting exoplanets with well defined parameters from the NASA exoplanet archive on 2026/01/16 to compare against NGTS-38 b. We used the planetary systems composite data table, containing 6071 entries, to avoid repeated entries of the same planet. First we eliminated dubious planets by setting the controversial flag to 0 and limited our sample to only transiting planets by setting the detected by transits flag to 1. We then similarly set the limit flags on radius, mass (eliminating M \sin i only constraints) and orbital periods to 0 to only allow constrained values on these parameters. We also only allowed systems with a constrained mass rather than M \sin i . This left us with 1434 planets. We then made a precision cut of 25% on mass, 10% on radius and 1% on orbital period, leaving 939 planets in the final comparison sample. We show NGTS-38 b plotted with this sample in Figure 9 . Of these, NGTS-38 b has the 16th longest period. (a) Period-radius (b) Period-eccentricity Figure 9 : Exoplanet population plots with NGTS-38 b overplotted as a black-bordered star. Orbital period on a base 10 logarithmic scale is plotted against planet radius (top) and eccentricity (bottom). The exoplanet population used is described in § 6.2.1 . Points on all of the plots are coloured by mass. 6.3 Interior Figure 10 : Planet mass in Jupiter masses (on a \log_{2} scale) plotted against planet radius in Jupiter radii for the transiting planet population described in § 6.2.1 . Planets are coloured by their equilibrium temperature. NGTS-38 b is shown as a black bordered star with errorbars. As can be seen in Figure 10 , NGTS-38 b appears to have a normal radius and density for a temperate planet of its mass, unaffected by inflation like many hotter planets. We used the planetary code completo ( 120 ) to estimate the heavy element content of NGTS-38 b. We built a grid of interior models assuming that planets are described with a layered model including a core with a mass comprised between 0 and 10 M ⊕ and an envelope coupled with a semi-grey atmospheric model. The heavy elements are modelled as water with the AQUA equation of state (EoS) from 77 and they are homogeneously mixed in the H/He envelope are described with EoS from 31 . The grid of evolution models was then coupled to a Bayesian inference model to retrieve the heavy element content that is compatible with the observed planetary mass, radius, equilibrium temperature, and age of the system. We find that the heavy element mass ( \rm M_{Z} ) is equal to \rm 110^{+90}_{-80} \rm M_{\oplus} . The overall stellar metallicity is approximated with a scaling law with the iron abundance \rm Z_{\star}=0.0142\times 10^{[Fe/H]} ( 10 ) . The planetary heavy element enrichment ( \rm Z_{p}/Z_{\star} , with \rm Z_{p}=M_{Z}/M_{P} ,) is equal to 2.4^{+2.0}_{-1.7} . This value is consistent with no metal enrichment of NGTS-38 b relative to its host star at 2\,\sigma . This result is in line with other high-mass ( \rm M>2\,M_{J} ) and weakly irradiated giant planets orbiting metal-rich stars, such as Kepler-1704 b ( 44 ) or TOI-2180 b ( 43 ) , which also show relatively low bulk metal enrichment. We note that the results of the interior structure modelling are dependent on several assumptions, in particular regarding the chosen interior structure and equations of state. The planetary metal enrichment can vary depending on these model assumptions but also on the interior modelling codes used to explore the planet’s composition. 7 Discussion 7.1 Alignment We do not find a photometric rotation period for TIC-65910228 / NGTS-38 to compare against the spectroscopic v\sin i_{\star} derived upper limit (see Section 4.4.1 ). This makes any constraints on the alignment of NGTS-38 b impossible without measuring the sky projected obliquity by observing the Rossiter-McLaughlin effect via a spectroscopic transit. If this was done, it would make NGTS-38 b the third longest period planet with a measured obliquity behind HIP 41378 d and f ( 72 ; 73 ) and one of only five such systems with periods beyond 100 d. The planet’s moderate eccentricity would also make it stand out among this sample, with the other four systems having either effectively circular or highly eccentric orbits (Source: NASA exoplanet archive; 2 ; 35 , Accessed 2026/01/16) . Of additional interest, is the fact that the effective temperature of the host star lies close to the Kraft break ( 107 ) , which has been linked to a bimodal distribution in observed giant planet obliquities ( 175 ; 6 ; 176 ; 164 ; 4 ; 102 ; 170 ) . Using equation 1 from 164 we estimate an RM amplitude of 14\pm 3 m s -1 for NGTS-38 b. This is within the detection limits of high precision radial velocity spectrographs such as the Echelle SPectrograph for Rocky Exoplanets and Stable Spectroscopic Observations ( 133 , ESPRESSO;) and potentially HARPS, albeit marginally. Although, the long transit duration (T {}_{14}= 14.88\pm 0.13 hrs) of NGTS-38 b will make such observations challenging from the ground. However, using multiple facilities across the globe it may be possible to obtain an effective observation baseline long enough to view the entire spectroscopic transit. This approach has already been effectively used for RM observations of the long-period planets in the HIP 41378 system ( 72 ; 73 ) . 7.2 Atmosphere As can be seen in Figure 11 , NGTS-38 b has one of the coolest equilibrium temperatures out of our sample of transiting exoplanets with well defined parameters (see Section 6.2.1 ). The equilibrium temperature of 458\pm 11 K is below the 500 K N 2 to NH 3 transition line ( 59 ) where the atmospheric nitrogen content switches between being dominated by diatomic Nitrogen to more complex molecular species. The planet’s host star also has one of the hottest effective temperatures of any planet with a temperature near-to / below this transition line, potentially leading to unique photochemistry in the upper atmosphere. Although, it should be noted that the equilibrium temperature at periastron ( 550\pm 14 K) is above this transition. It is also important to note that the chemical transition temperatures we quote from 59 are calculated for a 1 M J planet as this is the closest to the mass of NGTS-38 b of the three masses 59 calculate transition temperatures for. 59 note that the temperatures at which these transitions occur are also dependent on surface gravity, pressure and mixing ratios among other factors with a more massive planet like NGTS-38 b potentially seeing these transitions occur at higher temperatures. For a 10 M J planet the N 2 to NH 3 transition could instead occur at \sim 550 K instead of \sim 500 K for a 1 M J planet. In addition, the wide orbital separation of the planet means the rotation is unlikely to be tidally locked (see Section 6.1.1 ). The lack of a permanent day or night side will almost certainly have significant effects on the heat distribution and circulation dynamics of the planet. Similarly, the lack of tidal locking means the planetary rotation period could be much shorter than most hot-Jupiters, leading to further differences in atmospheric dynamics. For reviews into the atmospheric dynamics of giant planets, including rotational and tidal locking effects see: 135 and 154 . To test the observability of NGTS-38 b’s atmosphere, we calculate the Transmission Spectroscopy Metric ( 96 , TSM;) for NGTS-38 b, finding a low value of \sim 2.1 . The planet’s high mass results in a low atmospheric scale height and thus lower signal-to-noise compared to a less massive planet. This makes the planet a poor prospect for transmission spectroscopy follow-up. Because emission SNR is largely independent of planet mass we also investigate emission spectroscopy as an alternative approach to probing the planet’s atmosphere. Given the planet’s large orbital separation and relatively high transit impact parameter ( b_{\text{transit}} = 0.8540_{-0.010}^{+0.0090} ) a secondary eclipse would seem relatively unlikely. However, using equation 8 from 174 , we find a value of b_{\text{occultation}} = 0.59\pm 0.03 , meaning the planet is occulted by its host and emission spectroscopy may be possible. We calculate a value of the Emission Spectroscopy Metric ( 96 , ESM;) - although we choose to use the equilibrium temperature for NGTS-38 b instead of estimating a daytime temperature, since the planet is extremely unlikely to be tidally locked at its wide orbital separation. Using this method we find a value of \sim 22.5 , which, while modest, is relatively high for a planet with as low an equilibrium temperature as NGTS-38 b (see Figure 11 ). Figure 11 : Stellar effective temperature, T_{\text{eff}} , plotted against planetary equilibrium temperature, T_{\text{eq}} , for a sample of transiting planet systems described in § 6.2.1 . NGTS-38 b is overplotted as a black bordered star with errorbars sized to reflect the range of equilibrium temperatures throughout the planet’s orbit. Points are coloured according to their Emission Spectroscopy Metric ( 96 , ESM;) . Horizontal lines are plotted at 850, 700, 500 and 150 K to indicate the CO to CH 4 transition, KCl condensation, the N 2 to NH 3 transition and NH 3 condensation, respectively, for a 1 M J planet ( 59 ) . 7.3 Additional planetary companions Multiplicity is common in exoplanetary systems with 2574 out of 6071 (\sim 42\%) confirmed planets being in systems with other planets (Source: NASA exoplanet archive; 2 ; 35 , Accessed 2026/01/16) . A number of transiting warm Jupiter systems have already been found to host smaller inner companions ( 68 ; 9 , e.g. NGTS-11 b and c;) and warm-Jupiters appear to be more likely to have close planetary neighbours than their hot-Jupiter counterparts ( 86 ) . Given the wide orbital separation of NGTS-38 b it is fairly likely that additional interior planets could have stable orbits. With this in mind we performed a transit least squares search on the TESS dataset using the allesfitter package which acts as a wrapper for the transitleastsquares package ( 84 ) alongside wotan ( 83 ) for detrending. However, this search returned no convincing results. Additionally a generalised Lomb-Scargle periodogram of the RV residuals showed no significant peaks. Thus we detect no additional companions apart from NGTS-38 b, although this does not necessarily rule out the existence of any other companions in the system. 7.3.1 Sensitivity maps To ascertain the detection limits of the TESS data for NGTS-38 we made use of the Transit Investigation and Recoverability Application ( 143 , TIaRA ;) . TIaRA takes photometric data as inputs, specifically lightcurve timestamps, contamination values and measured photometric precision with the latter two in this instance being taken from the lightcurve fits file headers 1- CROWDSAP and CDPP2_0 , respectively. The sector 61 lightcurve lacked a recorded noise value so we instead used the sector 34 noise value in its place. TIaRA then uses the lightcurve properties to calculate the signal-to-noise and detection probability of generated signals producing a sensitivity map which we show with a smoothing function from 54 in Figure 12(a) . The resulting sensitivity map shows that planets smaller than 2-4 R ⊕ (depending on orbital period) are unlikely to be detectable in the TESS data, at least at periods shorter than \sim 100 days. This allows us to effectively rule out larger transiting companions than this, although, the aforementioned lack of any residual RV signal makes any large non-transiting companions additionally unlikely. Potential small inner transiting companions around NGTS-38 could be detected with the upcoming PLAnetary Transits and Oscillations of Stars ( 139 , PLATO ;) mission. This mission will observe a field spanning \sim 5% of the sky. For at least the first two years of its mission, PLATO will observe a field in the Southern Hemisphere, LOPS2 ( 125 ) . Unlike TESS which observes a target with one of its four cameras, PLATO has four sets of six cameras that overlap in the centre of its field. Hence a target can be observed with 6, 12, 18 or 24 cameras resulting in a higher precision of the photometric data. This and its longer observations enable PLATO to detect planets smaller than TESS ( 139 ; 54 ) . NGTS-38 lies within the PLATO LOPS2 field and will be monitored by PLATO with 6 cameras during its observation of the Southern field. These observations will not only enable us to determine the period and ephemeris more precisely but also to search for smaller inner companions. Following 54 , we apply TIaRA to 2 years of simulated PLATO data using the noise ( 26 ) reported in the PLATO Input Catalogue ( 119 ) . These simulations show that PLATO is predicted to detect potential inner companions down to sub 1 R ⊕ radii at short periods as we show in Figure 12(b) . (a) TESS (b) PLATO Figure 12 : TIaRA sensitivity maps for NGTS-38 in TESS (top) and PLATO (bottom) with NGTS-38 b overplotted as a white star with black borders. 7.4 Potential for the detection of exomoons and exorings We find no observational evidence of exomoons or exorings around NGTS-38 b. However, its large mass and wide orbital separation make it a potentially high value target in future searches for such circumplanetary systems. In the remainder of this section we assess the ability of the planet to form and retain moons and/or rings and the potential to detect them. 7.4.1 Formation and retention The planet’s current position is well interior to any primordial ice line of the system making the formation of icy moons or rings like those in the solar system unlikely if the planet formed in situ. If the planet formed beyond the ice line and migrated inwards any icy moons or rings would have to survive this migration. The theoretical lifespans of rings are too short to allow survival of such a ring system to the present. Moons could potentially survive such a migration, although dynamically violent migration mechanisms could potentially be more likely to disrupt any moons present ( 70 ) . Alternatively, moons or rings could be formed from the volatile poor silicate material found in the inner region of the disk. 13 find no meaningful upper limit on the mass of moons able to survive for planets orbiting their hosts beyond \sim 0.6 au. NGTS-38 b sits on the edge of this limit with its semi major axis of 0.695\pm 0.036 au, although its moderate eccentricity means the periastron distance falls below this limit. However, based on figure 2 of 13 this still allows the survivability of a \lessapprox 1 M ⊕ satellite. To further investigate the survivability and possible extent of circumplanetary systems around NGTS-38 b, we calculate the Hill sphere radius as follows: R_{\text{Hill}}=a\sqrt{\frac{\mbox{M\textsubscript{p}}}{3(\mbox{M\textsubscript{p}}+\mbox{M${}_{\star}$})}}, (1) finding a value of 147\pm 9 R J . This is much smaller than that of Jupiter (\sim 700\,\mbox{R\textsubscript{J}}) but still relatively large compared to most transiting exoplanets. 145 find that moons are stable out to \lessapprox 0.3 R Hill . For NGTS-38 b this is equivalent to a distance of 44.2\pm 2.8 R J . 49 similarly find an outer limit of stability for a moon on a circular prograde orbit of R_{\text{Hill}}\times 0.4895(1.0000-1.0305e) . For NGTS-38 b we find a value of 49.2\pm 3.2 R J , which we show plotted against host Gaia G-band magnitude in Figure 13 . Both these theoretical stable regions are greater than the separation between Jupiter and Callisto, the outermost Galilean moon. This means a Galilean moon system is tentatively within the stability limits of NGTS-38 b, however given the planet’s location interior to the ice line it is unclear whether such a system could form. We also calculate the Roche limit radius for NGTS-38 b as follows: d_{\text{Roche; p}}=2.44\mbox{R\textsubscript{p}}\left(\frac{\rho_{\text{b}}}{\rho_{\text{sat}}}\right)^{\frac{1}{3}}, (2) where \rho_{\text{sat}} is the density of a satellite, which we set to 2g cm -3 (similar to many asteroids and planetesimals). We calculate a value of 3.50\pm 0.28 R J , which is considerably larger than the limits for Jupiter (2.13 R J ) and Saturn (1.39 R J ). The Roche limit not only represents the innermost separation any potential exomoon could have without being disrupted but can also inform the extent of possible exorings. The ring systems around giant planets in the solar system tend to extend to 1-2 Roche limit radii, although proposed exoring candidates ( 98 , e.g. J1407 b;) are much larger and fill the entire hill sphere of their theoretical host planets. However, if such a ring system existed around NGTS-38 b it would have almost certainly already been detectable in the existing TESS and NGTS photometry. 7.4.2 Detectability of moons and rings 160 identify planets with periods >100 days transiting stars brighter than a Gaia G band magnitude of 11 as high priority targets for exomoon searches. NGTS-38 b comfortably meets both of these criteria (see Figure 13 ). We can refine this prioritisation further by only allowing planets where the outermost prograde circular orbit radius from 49 is larger than the separation between Jupiter and the outermost Galilean moon (Callisto). This leaves only 6 high priority targets, including NGTS-38 b (see Figure 13 ). The other five planets in this regime are HD-114082 b ( 178 ) , TIC-172900988 b ( 104 ) , TOI-2010 b ( 113 ) , TOI-2180 b ( 43 ) and TOI-4465 ( 57 ) . This makes NGTS-38 b one of the most exciting prospects for future exomoon surveys, although it should be noted that while the brightness of its host star will improve SNR, the large radius of the host will make detection of exomoons via the transit method more difficult. Jupiter’s largest moon; Ganymede, would have a transit depth of just over 14 ppm across the sun and only \sim 4 ppm for NGTS-38. We also considered the potential TTV signal of a Ganymede and an Earth mass moon on a circular orbit around NGTS-38 b at equal separation to that between Ganymede and Jupiter. We estimated the RMS amplitude of a TTV signal for both these cases using equation 3 from 100 , finding signals of 0.39\pm 0.04 s and 15.7\pm 1.5 s respectively. A signal this small would be extremely challenging to detect, making TTV exomoon detection for NGTS-38  b unlikely. It could be possible to detect exorings around the planet by searching for abnormalities in the planet’s transit shape or reflected light signals ( 12 ; 1 ; 3 ; 79 ) . We used the exorings software package ( 99 ; 98 ) to model an exoring with a radius equal to d_{\text{Roche}} around NGTS-38 b, and find that this model is indistinguishable from a ringless transiting planet in TESS . However, with the greater photometric precision available from instruments such as PLATO , JWST , or the ELT it may be possible to detect such a system. Alternatively a spectroscopic transit observation of the RM effect, which would already be of scientific interest for constraining the orbital obliquity of the system (see Section 7.1 ), could be used to detect an exoring system ( 46 ) . Figure 13 : Gaia G-band magnitude plotted against stable region radius from 49 for all planets in the sample described in § 6.2.1 . NGTS-38 b is plotted as a black bordered star. Points are coloured according to their orbital period. A grey vertical line is plotted at G=11 to represent the brightness cutoff from 160 . In addition, horizontal lines are plotted at 329 R J , 26.3 R J and 5.9 R J to represent the radius of the stable region for Jupiter, the distance between Jupiter and Callisto (the most widely separated Galilean moon), and the distance between Jupiter and the closest in Galilean moon; Io. 8 Summary and Conclusions TIC-65910228 b / NGTS-38 b was first detected from a single TESS transit in sector 33. We then initiated a campaign of spectroscopic and photometric followup that yielded an additional transit egress from NGTS and a full radial velocity solution from HARPS and CORALIE. The spectra obtained also allowed a characterisation of the host star, revealing it to have a temperature consistent with an F6V-F7V type star ( 6310\pm 129 K), and a low surface gravity ( \log g = 4.06\pm 0.02 ) resulting from a large radius ( 1.88\pm 0.08 R ⊙ ) compared to its mass ( 1.46\pm 0.09 M ⊙ ). This due to the enhanced metalicity ([Fe/H]= 0.33\pm 0.09 ‘dex’) of the star. A combined fit with the photometric and spectroscopic data confirmed the planet to be a 4.78_{-0.37}^{+0.40} M J , 1.081\pm 0.047 R J giant planet on a 180.52791\pm 0.00038 d orbit with a moderate eccentricity of 0.308\pm 0.011 . The planet’s wide orbital separation gives it a low equilibrium temperature compared to other transiting planets, varying between 550\pm 14 K and 400\pm 10 K throughout its eccentric orbit. On account of this cooler temperature, the atmosphere is likely to contain molecular nitrogen species not found in the atmospheres of hotter giant planets. In addition, this wide separation also makes NGTS-38 b one of the most likely transiting planets to host stable exomoon and/or ring systems which could be detectable with targeted observations from high precision facilities like PLATO , JWST or the ELT. While the planet is unsuitable for transmission spectroscopy (TSM \sim 2.9 ) with current generation instruments, there is a better chance of detecting an atmosphere through emission spectroscopy of the secondary eclipse (ESM \sim 23 ). However, perhaps the most pertinent additional observation of NGTS-38 b would be to constrain its orbital obliquity through spectroscopic transit observations. This will require extensive international collaboration to obtain longitudinal coverage from multiple facilities worldwide to observe the full \sim 15 hr transit and baseline. Although, with an estimated RM amplitude of 14\pm 3 m s -1 this should be possible. NGTS-38 b’s position within the LOPS2 field of the upcoming PLATO mission allows for the potential discovery of additional companions within the system. Either by directly detecting additional companions with transits too shallow and/or infrequent to be seen in current datasets or inferring outer companions through measuring transit timing variations of NGTS-38 b with its long baseline. Data Availability The TESS data is accessible via the MAST (Mikulski Archive for Space Telescopes) portal at https://mast.stsci.edu/portal/Mashup/Clients/Mast/Portal.html . Any code used for analysis or in producing the plots in this paper can be made available upon reasonable request to the author(s). Affiliations 1 Astrophysics Research Centre, School of Mathematics and Physics, Queen’s University Belfast, Belfast, BT7 1NN, UK 2 Leiden Observatory, Leiden University, P.O. Box 9513, 2300 RA Leiden, The Netherlands 3 Observatoire de Genève, Université de Genève, 51 Ch. des Maillettes, CH-1290 Sauverny, Switzerland 4 Department of Physics, University of Warwick, Gibbet Hill Road, Coventry CV4 7AL, UK 5 Centre for Exoplanets and Habitability, University of Warwick, Gibbet Hill Road, Coventry CV4 7AL, UK 6 Isaac Newton Group of Telescopes, Apartado de correos 321, E-38700 Santa Cruz de La Palma, Canary Islands; Spain 7 School of Physics & Astronomy, University of Birmingham, Edgbaston, Birmingham B15 2TT, UK 8 Center for Astrophysics | Harvard & Smithsonian, 60 Garden Street, Cambridge, MA 02138, USA 9 Department of Physics and Astronomy, The University of New Mexico, 210 Yale Blvd NE, Albuquerque, NM 87106, USA 10 Instituto de Astronomía, Universidad Católica del Norte, Angamos 0610, 1270709, Antofagasta, Chile 11 Instituto de Astrofísica de Canarias, c/ Vía Láctea s/n, 38205 La Laguna, Tenerife, Spain 12 Astrobiology Research Unit, Université de Liège, 19C Allée du 6 Août, 4000 Liège, Belgium 13 Department of Earth, Atmospheric and Planetary Science, Massachusetts Institute of Technology, 77 Massachusetts Avenue, Cambridge, MA 02139, USA 14 Astronomy Unit, Queen Mary University of London, Mile End Road, London E1 4NS, UK 15 School of Physics and Astronomy, University of Leicester, Leicester LE1 7RH, UK 16 El Sauce Observatory, Coquimbo Province, Chile 17 Rugby School, Lawrence Sheriff Street, Rugby, Warwickshire CV22 5EH, UK 18 Sapienza Università di Roma, Piazzale Aldo Moro, 5, 00185, Rome (RM), Italy 19 Campo Catino Astronomical Observatory, Regione Lazio, Guarcino (FR), 03010 Italy 20 INFN Sezione Roma1, Piazzale Aldo Moro, 2, 00185, Rome (RM), Italy 21 INAF OAC, Via della Scienza, 5, 09047, Selargius (CA), Italy 22 Instituto de Estudios Astrofísicos, Facultad de Ingeniería y Ciencias, Universidad Diego Portales, Av. Ejército Libertador 441, Santiago, Chile 23 Centro de Excelencia en Astrofísica y Tecnologías Afines (CATA), Camino El Observatorio 1515, Las Condes, Santiago, Chile 24 Columbia University, 550 W 120th Street, New York NY 10027, USA 25 National Research Council Canada, Herzberg Astronomy & Astrophysics Research Centre, 5071 West Saanich Road, Victoria, BC V9E 2E7, Canada 26 Subaru Telescope, National Astronomical Observatory of Japan, 650 North A‘ohoku Place, Hilo, HI 96720, USA 27 Departamento de Astrofísica, Universidad de La Laguna, 38206 La Laguna, Tenerife, Spain 28 Hamburger Sternwarte, Gojenbergsweg 112, 21029 Hamburg, Germany 29 Massachusetts Institute of Technology, Department of Physics and Kavli Institute for Astrophysics and Space Research, Cambridge, MA 02139, USA 30 Centre for Space Domain Awareness, University of Warwick, Gibbet Hill Road, Coventry CV4 7AL, UK 31 American Association of Variable Star Observers, 49 Bay State Road, Cambridge, MA 02138, USA 32 NASA Goddard Space Flight Center, 8800 Greenbelt Road, Greenbelt, MD 20771, USA 33 South African Astronomical Observatory, P.O Box 9, Observatory 7935, Cape Town, South Africa Appendix A NGTS photometric data In Table 4 we show the NGTS data for NGTS-38 separated into the runs shown in Table 1 . The table is truncated in print to only the first five rows for each run, however, full machine readable versions will be made available online. ActionID is an identification number given to each night of observations from each telescope. Camera is the ID of the telescope making the observations, in runs 1-3 this is only 809 while run 4 contains 809 and 810. Timestamps are listed in BJD. Relative Flux is the ‘raw’ flux output of the bsproc pipeline and is the target counts divided by the total master comparison counts. Normalised Flux is this relative flux median normalised for each run. The Sky background is given as the total number of electron counts for that exposure. The FWHM is measured (in pixels) by creating a super-sampled Point Spread Function (PSF) across the central third of the image. Table 4 : NGTS data. ActionID Camera Time Relative Flux Normalised Flux Airmass Sky Background FWHM BJD e - counts Pixels Run 1 354981 809 2460182.87854167 0.02155\pm 0.00026 1.002\pm 0.012 1.98486 30572.598 2.0288 354981 809 2460182.87877315 0.02177\pm 0.00026 1.012\pm 0.012 1.9807 30488.725 1.9034 354981 809 2460182.87892361 0.02162\pm 0.00026 1.005\pm 0.012 1.97804 30561.318 1.9016 354981 809 2460182.87907407 0.02140\pm 0.00026 0.995\pm 0.012 1.97533 30451.943 1.9763 354981 809 2460182.87922454 0.02146\pm 0.00026 0.998\pm 0.012 1.97268 30389.973 1.9915 … … … … … … … … Run 2 371970 809 2460417.47934028 0.01771\pm 0.00009 0.993\pm 0.005 1.04657 31509.871 1.7379 371970 809 2460417.47966435 0.01789\pm 0.00009 1.002\pm 0.005 1.04714 30847.371 1.7083 371970 809 2460417.47993056 0.01784\pm 0.00009 1.000\pm 0.005 1.0476 30342.18 1.7198 371970 809 2460417.48008102 0.01776\pm 0.00009 0.995\pm 0.005 1.04786 30145.863 1.7168 371970 809 2460417.48023148 0.01781\pm 0.00009 0.998\pm 0.005 1.04813 29873.92 1.7392 … … … … … … … … Run 3 380563 809 2460541.9097338 0.02013\pm 0.00026 1.018\pm 0.013 1.75903 67949.33 2.0486 380563 809 2460541.90997685 0.01997\pm 0.00026 1.010\pm 0.013 1.75561 67824.34 2.0401 380563 809 2460541.91012731 0.01986\pm 0.00025 1.004\pm 0.013 1.75361 67808.19 2.0626 380563 809 2460541.91027778 0.01980\pm 0.00025 1.002\pm 0.013 1.7516 68065.3 1.9862 380563 809 2460541.91042824 0.01966\pm 0.00025 0.994\pm 0.013 1.7496 67986.94 2.0328 … … … … … … … … Run 4 389750 809 2460637.63924769 0.01995\pm 0.00023 1.013\pm 0.011 1.9713 20280.932 1.9383 389750 809 2460637.63939815 0.01961\pm 0.00022 0.995\pm 0.011 1.96865 20225.166 1.8716 389750 809 2460637.63954861 0.01984\pm 0.00022 1.007\pm 0.011 1.96599 20302.121 1.9032 389750 809 2460637.63969907 0.01976\pm 0.00022 1.003\pm 0.011 1.96334 20340.297 1.9533 389750 809 2460637.63984954 0.01978\pm 0.00022 1.004\pm 0.011 1.96071 20321.646 1.9033 … … … … … … … … Run 5 422211 806 2461012.61271991 0.01583\pm 0.00021 1.017\pm 0.013 1.97804 56994.95 1.8726 422211 806 2461012.61297454 0.01530\pm 0.00020 0.983\pm 0.013 1.9736 56874.996 1.8558 422211 806 2461012.61311343 0.01542\pm 0.00020 0.991\pm 0.013 1.97096 56803.188 1.8159 422211 806 2461012.61326389 0.01541\pm 0.00020 0.990\pm 0.013 1.96832 56776.184 1.853 422211 806 2461012.61341435 0.01558\pm 0.00021 1.001\pm 0.013 1.96568 56766.62 1.8438 … … … … … … … … Appendix B Radial Velocity data Here we present the radial velocity timeseries data from both CORALIE (Table 5 ) and HARPS (Table 6 ). These tables will be available online in a machine readable format. For both datasets Time is the Time in TBJD (i.e. BJD-2457000) rounded to 3 decimal places and RV is the radial velocity in km s -1 . For HARPS we list the exposure time T exp in minutes for each spectrum whilst we omit this for CORALIE as every spectrum was taken with an exposure time of 20 minutes. For CORALIE SNR 62 is the signal to noise ratio of spectral order 62 and for HARPS SNR 64 is the signal to noise ratio of spectral order 64. For both datasets, ‘Airmass’ is the airmass at the midpoint of the exposure for each spectrum. FWHM and BIS are the Full Width Half Maximum and Bisector Inverse Slopes of the Cross Correlation Function used to determine the RVs of each spectrum in km s -1 and m s -1 respectively. For Both datasets we also include the Contrast as well as the Hydrogen alpha, Sodium and Calcium indexes as dimensionless quantities. We list the Barycentric Radial Velocity correction (Berv) in m s -1 for both datasets. For HARPS we also include the \log\,\mathrm{R}^{\prime}_{\mathrm{HK}} value calculated for each spectra. Table 5 : CORALIE data. Time RV SNR 62 Airmass FWHM BIS Contrast H \alpha index Na index Ca index Berv (TBJD) km s -1 km s -1 m s -1 m s -1 CORALIE 14 3050.559 20.15\pm 0.03 23.3 1.29 12.87\pm 0.02 25\pm 39 24.71 0.180\pm 0.004 0.372\pm 0.005 0.026\pm 0.008 -18.87 3289.700 20.04\pm 0.02 31.8 1.11 12.84\pm 0.02 2\pm 22 24.77 0.184\pm 0.003 0.372\pm 0.003 0.130\pm 0.004 11.72 3310.870 19.99\pm 0.01 36.1 1.41 12.88\pm 0.02 -20\pm 20 24.47 0.178\pm 0.003 0.358\pm 0.003 0.142\pm 0.004 5.05 3367.677 19.91\pm 0.02 34.5 1.20 12.91\pm 0.02 64\pm 23 24.38 0.185\pm 0.003 0.363\pm 0.003 0.136\pm 0.005 -12.48 3381.711 19.99\pm 0.02 35.3 1.74 12.95\pm 0.02 79\pm 23 24.33 0.183\pm 0.003 0.360\pm 0.003 0.131\pm 0.006 -15.63 3394.672 20.11\pm 0.01 37.9 1.76 12.86\pm 0.02 -10\pm 21 24.39 0.178\pm 0.003 0.354\pm 0.003 0.135\pm 0.005 -17.60 3397.533 20.12\pm 0.02 31.6 1.03 12.83\pm 0.02 23\pm 26 24.47 0.170\pm 0.003 0.350\pm 0.003 0.199\pm 0.006 -17.71 3401.571 20.17\pm 0.01 40.4 1.15 12.95\pm 0.02 21\pm 20 24.36 0.179\pm 0.003 0.354\pm 0.003 0.125\pm 0.004 -18.19 3404.563 20.16\pm 0.01 39.2 1.16 12.87\pm 0.02 -27\pm 20 24.44 0.187\pm 0.003 0.352\pm 0.003 0.107\pm 0.004 -18.42 3409.578 20.19\pm 0.03 23.5 1.28 12.90\pm 0.02 -37\pm 40 24.51 0.180\pm 0.004 0.349\pm 0.004 0.127\pm 0.009 -18.74 3424.509 20.18\pm 0.02 34.3 1.16 12.90\pm 0.02 -10\pm 25 24.26 0.188\pm 0.003 0.359\pm 0.003 0.132\pm 0.005 -18.66 3428.525 20.15\pm 0.02 30.5 1.29 12.84\pm 0.02 43\pm 29 24.55 0.176\pm 0.003 0.344\pm 0.004 0.135\pm 0.006 -18.50 3435.482 20.15\pm 0.02 35.2 1.18 12.86\pm 0.02 29\pm 23 24.52 0.178\pm 0.003 0.365\pm 0.003 0.144\pm 0.005 -17.91 CORALIE 24 3605.815 20.15\pm 0.02 24.9 1.17 12.92\pm 0.02 65\pm 32 24.54 0.192\pm 0.004 0.382\pm 0.004 0.128\pm 0.006 18.93 3614.759 20.14\pm 0.03 19.6 1.32 12.86\pm 0.02 -7\pm 43 24.68 0.185\pm 0.005 0.376\pm 0.005 0.162\pm 0.010 18.64 3621.745 20.10\pm 0.02 31.1 1.29 12.78\pm 0.02 -11\pm 24 24.57 0.191\pm 0.003 0.355\pm 0.003 0.162\pm 0.005 18.03 3630.687 20.10\pm 0.02 25.9 1.55 12.89\pm 0.02 65\pm 29 24.72 0.182\pm 0.004 0.363\pm 0.004 0.103\pm 0.007 16.93 3637.713 20.10\pm 0.02 31.3 1.23 12.87\pm 0.02 -1\pm 23 24.55 0.190\pm 0.003 0.370\pm 0.003 0.150\pm 0.005 15.68 3659.727 19.98\pm 0.01 33.7 1.02 12.84\pm 0.02 -28\pm 21 24.60 0.184\pm 0.003 0.367\pm 0.003 0.137\pm 0.004 10.35 3666.641 20.01\pm 0.02 27.9 1.20 12.81\pm 0.02 -18\pm 26 24.72 0.190\pm 0.004 0.355\pm 0.004 0.135\pm 0.005 8.50 3677.697 19.99\pm 0.02 27.8 1.00 12.84\pm 0.02 -0\pm 27 24.66 0.183\pm 0.004 0.348\pm 0.004 0.150\pm 0.006 4.88 3684.604 19.95\pm 0.02 26.3 1.16 12.81\pm 0.02 0\pm 28 24.72 0.183\pm 0.004 0.346\pm 0.004 0.140\pm 0.006 2.78 3694.556 19.97\pm 0.02 29.5 1.25 12.89\pm 0.02 28\pm 25 24.62 0.174\pm 0.003 0.352\pm 0.003 0.138\pm 0.005 -0.47 3703.620 19.90\pm 0.01 34.2 1.01 12.82\pm 0.02 42\pm 21 24.62 0.181\pm 0.003 0.349\pm 0.003 0.123\pm 0.004 -3.63 3710.719 19.91\pm 0.02 29.3 1.13 12.83\pm 0.02 21\pm 26 24.60 0.178\pm 0.003 0.352\pm 0.004 0.171\pm 0.006 -6.16 3715.699 19.92\pm 0.04 15.3 1.11 12.83\pm 0.02 65\pm 63 24.44 0.183\pm 0.006 0.349\pm 0.006 0.131\pm 0.015 -7.67 3722.634 19.92\pm 0.02 28.3 1.02 12.85\pm 0.02 -54\pm 26 24.57 0.179\pm 0.004 0.350\pm 0.004 0.121\pm 0.005 -9.60 3727.647 19.91\pm 0.02 31.0 1.06 12.89\pm 0.02 -10\pm 22 24.60 0.175\pm 0.003 0.348\pm 0.003 0.148\pm 0.004 -11.03 3734.658 19.93\pm 0.02 27.3 1.15 12.92\pm 0.02 58\pm 27 24.59 0.184\pm 0.004 0.352\pm 0.004 0.132\pm 0.006 -12.88 3747.587 20.02\pm 0.02 30.4 1.05 12.82\pm 0.02 61\pm 24 24.57 0.181\pm 0.003 0.359\pm 0.003 0.150\pm 0.005 -15.54 3754.633 20.08\pm 0.01 34.9 1.28 12.92\pm 0.02 51\pm 20 24.56 0.184\pm 0.003 0.352\pm 0.003 0.142\pm 0.004 -16.82 3760.643 20.17\pm 0.02 23.3 1.47 12.87\pm 0.02 22\pm 33 24.62 0.178\pm 0.004 0.362\pm 0.004 0.127\pm 0.008 -17.66 3765.528 20.17\pm 0.02 22.5 1.04 12.93\pm 0.02 43\pm 35 24.58 0.184\pm 0.004 0.360\pm 0.005 0.085\pm 0.007 -17.99 3773.532 20.18\pm 0.02 33.4 1.09 12.88\pm 0.02 13\pm 22 24.57 0.186\pm 0.003 0.371\pm 0.003 0.133\pm 0.005 -18.58 3777.505 20.16\pm 0.02 28.4 1.05 12.86\pm 0.02 29\pm 27 24.58 0.193\pm 0.004 0.362\pm 0.004 0.151\pm 0.006 -18.69 3782.499 20.18\pm 0.02 32.9 1.07 12.89\pm 0.02 6\pm 23 24.47 0.182\pm 0.003 0.359\pm 0.003 0.115\pm 0.005 -18.76 3784.553 20.16\pm 0.03 22.8 1.31 12.87\pm 0.02 -13\pm 37 24.53 0.183\pm 0.004 0.372\pm 0.005 0.085\pm 0.008 -18.85 3789.497 20.16\pm 0.02 26.6 1.12 12.91\pm 0.02 -16\pm 30 24.56 0.189\pm 0.004 0.359\pm 0.004 0.154\pm 0.007 -18.66 3795.509 20.13\pm 0.02 29.1 1.23 12.90\pm 0.02 -17\pm 27 24.62 0.182\pm 0.004 0.360\pm 0.004 0.127\pm 0.006 -18.37 3801.496 20.14\pm 0.02 28.5 1.25 12.89\pm 0.02 -2\pm 28 24.48 0.181\pm 0.004 0.367\pm 0.004 0.122\pm 0.006 -17.85 3808.494 20.13\pm 0.02 28.6 1.35 12.90\pm 0.02 7\pm 29 24.18 0.189\pm 0.004 0.362\pm 0.004 0.181\pm 0.007 -17.03 3818.465 20.09\pm 0.02 33.1 1.35 12.87\pm 0.02 10\pm 24 24.40 0.182\pm 0.003 0.353\pm 0.003 0.153\pm 0.006 -15.46 3826.444 20.05\pm 0.02 30.2 1.34 12.81\pm 0.02 68\pm 28 24.23 0.184\pm 0.003 0.360\pm 0.004 0.148\pm 0.007 -13.88 3834.458 20.03\pm 0.02 27.9 1.66 12.90\pm 0.02 25\pm 32 24.24 0.186\pm 0.004 0.363\pm 0.004 0.094\pm 0.009 -12.09 3949.836 20.13\pm 0.02 23.2 1.34 12.83\pm 0.02 20\pm 33 24.61 0.183\pm 0.004 0.369\pm 0.004 0.215\pm 0.008 18.06 3957.846 20.18\pm 0.02 29.0 1.17 12.86\pm 0.02 -20\pm 25 24.60 0.185\pm 0.004 0.375\pm 0.004 0.126\pm 0.005 18.64 3964.873 20.16\pm 0.02 29.2 1.05 12.82\pm 0.02 -17\pm 25 24.56 0.191\pm 0.004 0.368\pm 0.004 0.131\pm 0.005 18.84 Table 6 : HARPS data. Time RV T {}_{\text{exp}} SNR 64 Airmass FWHM BIS Contrast H \alpha index Na index Ca index Berv \log\,\mathrm{R}^{\prime}_{\mathrm{HK}} (TBJD) km s -1 (mins) km s -1 m s -1 m s -1 3299.723 19.916\pm 0.003 25 75.5 1.02 12.56\pm 0.01 2\pm 7 37.55 0.1926\pm 0.0002 0.3924\pm 0.0002 0.1084\pm 0.0004 8.82 -5.012\pm 0.003 3325.788 19.838\pm 0.004 25 63.6 1.19 12.58\pm 0.01 -13\pm 8 37.53 0.1951\pm 0.0003 0.3912\pm 0.0002 0.1059\pm 0.0006 0.22 -5.032\pm 0.005 3328.702 19.848\pm 0.004 25 56.8 1.01 12.55\pm 0.01 3\pm 9 37.51 0.1976\pm 0.0003 0.3893\pm 0.0002 0.1077\pm 0.0006 -0.59 -5.017\pm 0.005 3331.723 19.828\pm 0.005 25 51.4 1.05 12.54\pm 0.01 -8\pm 10 37.49 0.1981\pm 0.0004 0.3896\pm 0.0003 0.1072\pm 0.0007 -1.65 -5.022\pm 0.006 3342.747 19.817\pm 0.004 25 75.9 1.21 12.59\pm 0.01 -15\pm 7 37.54 0.1911\pm 0.0002 0.3890\pm 0.0002 0.1067\pm 0.0005 -5.33 -5.026\pm 0.004 3346.790 19.809\pm 0.005 25 58.9 1.59 12.54\pm 0.01 -6\pm 10 37.56 0.1937\pm 0.0003 0.3894\pm 0.0002 0.0993\pm 0.0010 -6.68 -5.090\pm 0.009 3350.712 19.804\pm 0.003 25 78.4 1.16 12.58\pm 0.01 5\pm 7 37.48 0.1932\pm 0.0002 0.3875\pm 0.0002 0.1056\pm 0.0005 -7.78 -5.035\pm 0.004 3375.668 19.835\pm 0.004 20 65.1 1.27 12.56\pm 0.01 3\pm 8 37.55 0.1938\pm 0.0002 0.3902\pm 0.0002 0.1067\pm 0.0006 -14.34 -5.026\pm 0.005 3380.527 19.863\pm 0.004 20 53.2 1.00 12.57\pm 0.01 -1\pm 9 37.57 0.1966\pm 0.0003 0.3917\pm 0.0003 0.1101\pm 0.0005 -15.07 -4.999\pm 0.004 3598.796 20.071\pm 0.005 20 46.1 1.36 12.54\pm 0.01 13\pm 11 37.59 0.2033\pm 0.0004 0.3990\pm 0.0003 0.0924\pm 0.0008 18.95 -5.161\pm 0.009 3601.803 20.057\pm 0.005 20 44.9 1.26 12.54\pm 0.01 -13\pm 11 37.62 0.2019\pm 0.0004 0.4011\pm 0.0003 0.0906\pm 0.0008 18.98 -5.182\pm 0.009 3614.774 20.044\pm 0.006 20 42.7 1.24 12.58\pm 0.01 -5\pm 12 37.58 0.2035\pm 0.0004 0.3991\pm 0.0004 0.0964\pm 0.0010 18.61 -5.119\pm 0.010 3660.672 19.918\pm 0.004 20 55.0 1.15 12.58\pm 0.01 -1\pm 9 37.56 0.1986\pm 0.0003 0.3940\pm 0.0003 0.0937\pm 0.0006 10.19 -5.147\pm 0.007 3702.720 19.830\pm 0.005 20 53.0 1.07 12.58\pm 0.01 -0\pm 9 37.58 0.1951\pm 0.0003 0.3994\pm 0.0002 0.1025\pm 0.0008 -3.55 -5.061\pm 0.007 3731.658 19.826\pm 0.005 20 52.8 1.13 12.57\pm 0.01 11\pm 9 37.58 0.1959\pm 0.0003 0.4010\pm 0.0003 0.0945\pm 0.0007 -12.12 -5.138\pm 0.008 3737.647 19.847\pm 0.005 20 49.7 1.14 12.56\pm 0.01 12\pm 10 37.56 0.1989\pm 0.0003 0.3977\pm 0.0003 0.0945\pm 0.0009 -13.57 -5.139\pm 0.009 3747.656 19.925\pm 0.004 20 59.0 1.31 12.58\pm 0.01 -12\pm 9 37.45 0.1930\pm 0.0003 0.4009\pm 0.0002 0.0987\pm 0.0008 -15.67 -5.097\pm 0.007 3753.588 19.983\pm 0.005 20 52.2 1.10 12.57\pm 0.01 -14\pm 9 37.46 0.1983\pm 0.0003 0.4067\pm 0.0002 0.0833\pm 0.0007 -16.57 -5.277\pm 0.010 3765.630 20.063\pm 0.005 20 49.9 1.49 12.53\pm 0.01 -8\pm 11 37.51 0.1997\pm 0.0004 0.4040\pm 0.0003 0.0872\pm 0.0010 -18.17 -5.223\pm 0.013 3784.578 20.063\pm 0.008 20 35.3 1.49 12.55\pm 0.02 11\pm 16 37.67 0.2046\pm 0.0006 0.4073\pm 0.0004 0.0643\pm 0.0018 -18.89 -5.716\pm 0.071 3800.528 20.029\pm 0.004 20 63.2 1.45 12.58\pm 0.01 2\pm 8 37.46 0.1925\pm 0.0002 0.4099\pm 0.0002 0.0921\pm 0.0007 -17.99 -5.165\pm 0.008 3801.463 20.034\pm 0.004 20 57.0 1.13 12.55\pm 0.01 -11\pm 8 37.53 0.1934\pm 0.0003 0.4078\pm 0.0002 0.0900\pm 0.0006 -17.79 -5.189\pm 0.007 4045.820 19.862\pm 0.005 20 56.0 1.20 12.61\pm 0.01 -2\pm 10 37.63 0.2000\pm 0.0003 0.3794\pm 0.0003 0.0965\pm 0.0008 3.69 -5.118\pm 0.008 Appendix C allesfitter results tables Here we present the full set of fitted (Table 7 ) and derived (Table 8 ) parameters of NGTS-38 b produced by allesfitter (see Section 5 ). Some of these parameters are repeated in Table 3 . These will be available online in a machine readable format. Table 7 : allesfitter global fitted values and priors. Parameter Initial guess Prior Fitted value units Astrophysical parameters R_{b}/R_{\star} 0.06 \mathcal{U}\left(0,0.1\right) 0.05909\pm 0.00061 (R_{\star}+R_{b})/a_{b} 0.014 \mathcal{U}\left(0,0.05\right) 0.01332\pm 0.00037 \cos{i_{b}} 0.009 \mathcal{U}\left(0,1\right) 0.00825\pm 0.00033 T_{0;b} 2459209.228 \mathcal{U}\left(2459208,2459210\right) 2459209.2286\pm 0.0031 BJD P_{b} 180.5 \mathcal{U}\left(179,181\right) 180.52791\pm 0.00038 days \sqrt{e_{b}}\cos{\omega_{b}} 0.1 \mathcal{U}\left(-1,1\right) 0.0783\pm 0.0058 \sqrt{e_{b}}\sin{\omega_{b}} -0.5 \mathcal{U}\left(-1,1\right) -0.5493\pm 0.010 K_{b} 0.14 \mathcal{U}\left(0,0.5\right) 0.1372\pm 0.0018 km s -1 Instrumental limb darkening parameters q_{1;\mathrm{TESS}} 0.278 \mathcal{N}\left(0.278,0.002\right) 0.2782\pm 0.0019 q_{2;\mathrm{TESS}} 0.375 \mathcal{N}\left(0.375,0.025\right) 0.385\pm 0.024 q_{1;\mathrm{NGTS}} 0.337 \mathcal{N}\left(0.337,0.003\right) 0.3368\pm 0.0027 q_{2;\mathrm{NGTS}} 0.385 \mathcal{N}\left(0.385,0.025\right) 0.385_{-0.024}^{+0.023} q_{1;\mathrm{LCO}} 0.283 \mathcal{N}\left(0.283,0.003\right) 0.2827\pm 0.0029 q_{2;\mathrm{LCO}} 0.379 \mathcal{N}\left(0.379,0.026\right) 0.373_{-0.025}^{+0.023} q_{1;\mathrm{ElSauce}} 0.376 \mathcal{N}\left(0.376,0.009\right) 0.3750\pm 0.0085 q_{2;\mathrm{ElSauce}} 0.388 \mathcal{N}\left(0.388,0.073\right) 0.381\pm 0.067 q_{1;\mathrm{DSC0.4m}} 0.337 \mathcal{N}\left(0.337,0.009\right) 0.3376\pm 0.0086 q_{2;\mathrm{DSC0.4m}} 0.384 \mathcal{N}\left(0.384,0.076\right) 0.372\pm 0.071 q_{1;\mathrm{OACC-CAO}} 0.269 \mathcal{N}\left(0.269,0.008\right) 0.2689\pm 0.0076 q_{2;\mathrm{OACC-CAO}} 0.376 \mathcal{N}\left(0.376,0.079\right) 0.382\pm 0.073 Instrumental white noise parameters \ln{\sigma_{\mathrm{TESS}}} -6.0 \mathcal{U}\left(-14.0,0.0\right) -7.750\pm 0.011 \ln{\mathrm{rel.flux.}} \ln{\sigma_{\mathrm{NGTS}}} -6.0 \mathcal{U}\left(-14.0,0.0\right) -4.900\pm 0.011 \ln{\mathrm{rel.flux.}} \ln{\sigma_{\mathrm{LCO-CTIO(1m)}}} -6.0 \mathcal{U}\left(-14.0,0.0\right) -6.250\pm 0.020 \ln{\mathrm{rel.flux.}} \ln{\sigma_{\mathrm{LCO-CTIO(0.35m)}}} -6.0 \mathcal{U}\left(-14.0,0.0\right) -5.749\pm 0.026 \ln{\mathrm{rel.flux.}} \ln{\sigma_{\mathrm{LCO-SSO(1m)}}} -6.0 \mathcal{U}\left(-14.0,0.0\right) -6.668\pm 0.022 \ln{\mathrm{rel.flux.}} \ln{\sigma_{\mathrm{LCO-SSO(0.35m)}}} -6.0 \mathcal{U}\left(-14.0,0.0\right) -5.966\pm 0.020 \ln{\mathrm{rel.flux.}} \ln{\sigma_{\mathrm{ElSauce}}} -6.0 \mathcal{U}\left(-14.0,0.0\right) -5.134\pm 0.025 \ln{\mathrm{rel.flux.}} \ln{\sigma_{\mathrm{DSC0.4m}}} -6.0 \mathcal{U}\left(-14.0,0.0\right) -6.251\pm 0.063 \ln{\mathrm{rel.flux.}} \ln{\sigma_{\mathrm{OACC-CAO}}} -6.0 \mathcal{U}\left(-14.0,0.0\right) -6.153_{-0.046}^{+0.048} \ln{\mathrm{rel.flux.}} \ln{\sigma_{\mathrm{CORALIE14}}} -5.0 \mathcal{U}\left(-7.6,-4.6\right) -6.32_{-0.81}^{+0.91} \ln km s -1 \ln{\sigma_{\mathrm{CORALIE24}}} -5.0 \mathcal{U}\left(-7.6,-4.6\right) -6.21_{-0.90}^{+0.96} \ln km s -1 \ln{\sigma_{\mathrm{HARPS}}} -5.0 \mathcal{U}\left(-7.6,-4.6\right) -5.40_{-0.39}^{+0.30} \ln km s -1 Instrumental red noise / offset parameters offset CORALIE14 20.03 \mathcal{U}\left(19.95,20.05\right) 20.0357\pm 0.0046 km s -1 offset CORALIE24 20.03 \mathcal{U}\left(19.95,20.05\right) 20.0315\pm 0.0031 km s -1 offset HARPS 19.93 \mathcal{U}\left(19.9,20.0\right) 19.9306\pm 0.0014 km s -1 offset NGTS -0.0005 \mathcal{U}\left(-0.001,0.001\right) -0.00034\pm 0.00011 offset LCO-CTIO (1m) 0.0 \mathcal{U}\left(-0.001,0.001\right) -0.000327\pm 0.000068 offset LCO-CTIO (0.35m) 0.0 \mathcal{U}\left(-0.001,0.001\right) -0.00032\pm 0.00012 offset LCO-SSO (1m) 0.0 \mathcal{U}\left(-0.001,0.001\right) 0.000025\pm 0.000042 offset LCO-SSO (0.35m) 0.0 \mathcal{U}\left(-0.001,0.001\right) -0.000048_{-0.000071}^{+0.000076} offset ElSauce 0.0 \mathcal{U}\left(-0.001,0.001\right) -0.00057\pm 0.00020 offset DSC0.4m 0.0 \mathcal{U}\left(-0.001,0.001\right) -0.00066_{-0.00017}^{+0.00018} offset OACC-CAO 0.0 \mathcal{U}\left(-0.001,0.001\right) -0.00065\pm 0.00015 Table 8 : allesfitter global model derived values. Parameter Value Unit Astrophysical parameters R_{\star}/a_{\mathrm{b}} 0.01258\pm 0.00035 a_{\mathrm{b}}/R_{\star} 79.5_{-2.1}^{+2.2} R_{\mathrm{b}}/a_{\mathrm{b}} 0.000743\pm 0.000026 R_{\mathrm{b}} 12.12\pm 0.53 R ⊕ R_{\mathrm{b}} 1.081\pm 0.047 R J a_{\mathrm{b}} 149.5\pm 7.7 R ⊙ a_{\mathrm{b}} 0.695\pm 0.036 au i_{\mathrm{b}} 89.528\pm 0.019 deg e_{\mathrm{b}} 0.308\pm 0.011 \omega_{\mathrm{b}} 278.11\pm 0.70 deg q_{\mathrm{b}} 0.00313_{-0.00016}^{+0.00017} M_{\mathrm{b}} 1520_{-120}^{+130} M ⊕ M_{\mathrm{b}} 4.78_{-0.37}^{+0.40} M J M_{\mathrm{b}} 0.00456_{-0.00036}^{+0.00038} M ⊙ b_{\mathrm{tra;b}} 0.8540_{-0.010}^{+0.0090} T_{\mathrm{tot;b}} 14.88\pm 0.13 hours T_{\mathrm{full;b}} 9.38_{-0.29}^{+0.27} hours \rho_{\mathrm{\star}} 0.292_{-0.022}^{+0.025} cgs \rho_{\mathrm{b}} 4.69_{-0.78}^{+0.95} cgs g_{\mathrm{b}} 9510_{-630}^{+700} cgs T_{\mathrm{eq;b}} 458\pm 11 K Transit depth per Instrument \delta_{\mathrm{tr;b;{\it TESS}}} 3.200\pm 0.051 parts per thousand (ppt) \delta_{\mathrm{tr;b;NGTS}} 3.169\pm 0.050 parts per thousand (ppt) \delta_{\mathrm{tr;b;LCO-SSO1m}} 3.206\pm 0.052 parts per thousand (ppt) \delta_{\mathrm{tr;b;LCO-SSO0.35m}} 3.207_{-0.051}^{+0.054} parts per thousand (ppt) \delta_{\mathrm{tr;b;LCO-CTIO1m}} 3.203_{-0.045}^{+0.056} parts per thousand (ppt) \delta_{\mathrm{tr;b;LCO-CTIO0.35m}} 3.205\pm 0.049 parts per thousand (ppt) \delta_{\mathrm{tr;b;ElSauce}} 3.154\pm 0.062 parts per thousand (ppt) \delta_{\mathrm{tr;b;DSC0.4m}} 3.175\pm 0.065 parts per thousand (ppt) \delta_{\mathrm{tr;b;OACC-CAO}} 3.212\pm 0.060 parts per thousand (ppt) Classical limb darkening parameters per instrument u_{\mathrm{1;{\it TESS}}} 0.406\pm 0.026 u_{\mathrm{2;{\it TESS}}} 0.121\pm 0.025 u_{\mathrm{1;NGTS}} 0.447\pm 0.028 u_{\mathrm{2;NGTS}} 0.134\pm 0.028 u_{\mathrm{1;LCOGT}} 0.397_{-0.026}^{+0.024} u_{\mathrm{2;LCOGT}} 0.135_{-0.024}^{+0.026} u_{\mathrm{1;ElSauce}} 0.466\pm 0.083 u_{\mathrm{2;ElSauce}} 0.146\pm 0.083 u_{\mathrm{1;DSC0.4m}} 0.432\pm 0.083 u_{\mathrm{2;DSC0.4m}} 0.149\pm 0.083 u_{\mathrm{1;OACC-CAO}} 0.397_{-0.077}^{+0.073} u_{\mathrm{2;OACC-CAO}} 0.122\pm 0.076