diff --git a/joss.06104/10.21105.joss.06104.crossref.xml b/joss.06104/10.21105.joss.06104.crossref.xml new file mode 100644 index 0000000000..695144a6c0 --- /dev/null +++ b/joss.06104/10.21105.joss.06104.crossref.xml @@ -0,0 +1,634 @@ + + + + 20240810110440-04af05f4317957dbf6fa07f5ca5c97402c02495f + 20240810110440 + + JOSS Admin + admin@theoj.org + + The Open Journal + + + + + Journal of Open Source Software + JOSS + 2475-9066 + + 10.21105/joss + https://joss.theoj.org + + + + + 08 + 2024 + + + 9 + + 100 + + + + cortecs: A Python package for compressing +opacities + + + + Arjun B. + Savel + https://orcid.org/0000-0002-2454-768X + + + Megan + Bedell + https://orcid.org/0000-0001-9907-7742 + + + Eliza M.-R. + Kempton + https://orcid.org/0000-0002-1337-9051 + + + + 08 + 10 + 2024 + + + 6104 + + + 10.21105/joss.06104 + + + http://creativecommons.org/licenses/by/4.0/ + http://creativecommons.org/licenses/by/4.0/ + http://creativecommons.org/licenses/by/4.0/ + + + + Software archive + 10.5281/zenodo.13208759 + + + GitHub review issue + https://github.com/openjournals/joss-reviews/issues/6104 + + + + 10.21105/joss.06104 + https://joss.theoj.org/papers/10.21105/joss.06104 + + + https://joss.theoj.org/papers/10.21105/joss.06104.pdf + + + + + + Including all the lines: Data releases for +spectra and opacities + Kurucz + Canadian Journal of Physics + 9 + 95 + 10.1139/cjp-2016-0794 + 2017 + Kurucz, R. L. (2017). Including all +the lines: Data releases for spectra and opacities. Canadian Journal of +Physics, 95(9), 825–827. +https://doi.org/10.1139/cjp-2016-0794 + + + Atmospheric dynamics of hot giant planets and +brown dwarfs + Showman + Space Science Reviews + 216 + 10.1007/s11214-020-00758-8 + 2020 + Showman, A. P., Tan, X., & +Parmentier, V. (2020). Atmospheric dynamics of hot giant planets and +brown dwarfs. Space Science Reviews, 216, 1–83. +https://doi.org/10.1007/s11214-020-00758-8 + + + A comparison of exoplanet spectroscopic +retrieval tools + Barstow + Monthly Notices of the Royal Astronomical +Society + 4 + 493 + 10.1093/mnras/staa548 + 2020 + Barstow, J. K., Changeat, Q., +Garland, R., Line, M. R., Rocchetto, M., & Waldmann, I. P. (2020). A +comparison of exoplanet spectroscopic retrieval tools. Monthly Notices +of the Royal Astronomical Society, 493(4), 4884–4909. +https://doi.org/10.1093/mnras/staa548 + + + A solar c/o and sub-solar metallicity in a +hot jupiter atmosphere + Line + Nature + 7882 + 598 + 10.1038/s41586-021-03912-6 + 2021 + Line, M. R., Brogi, M., Bean, J. L., +Gandhi, S., Zalesky, J., Parmentier, V., Smith, P., Mace, G. N., +Mansfield, M., Kempton, E. M.-R., & others. (2021). A solar c/o and +sub-solar metallicity in a hot jupiter atmosphere. Nature, 598(7882), +580–584. +https://doi.org/10.1038/s41586-021-03912-6 + + + The ExoMol database: Molecular line lists for +exoplanet and other hot atmospheres + Tennyson + Journal of Molecular +Spectroscopy + 327 + 10.1016/j.jms.2016.05.002 + 2016 + Tennyson, J., Yurchenko, S. N., +Al-Refaie, A. F., Barton, E. J., Chubb, K. L., Coles, P. A., +Diamantopoulou, S., Gorman, M. N., Hill, C., Lam, A. Z., & others. +(2016). The ExoMol database: Molecular line lists for exoplanet and +other hot atmospheres. Journal of Molecular Spectroscopy, 327, 73–94. +https://doi.org/10.1016/j.jms.2016.05.002 + + + Confirmation of water absorption in the +thermal emission spectrum of the hot jupiter WASP-77Ab with +HST/WFC3 + Mansfield + The Astronomical Journal + 6 + 163 + 10.3847/1538-3881/ac658f + 2022 + Mansfield, M., Wiser, L., Stevenson, +K. B., Smith, P., Line, M. R., Bean, J. L., Fortney, J. J., Parmentier, +V., Kempton, E. M.-R., Arcangeli, J., & others. (2022). Confirmation +of water absorption in the thermal emission spectrum of the hot jupiter +WASP-77Ab with HST/WFC3. The Astronomical Journal, 163(6), 261. +https://doi.org/10.3847/1538-3881/ac658f + + + Confirmation of Subsolar Metallicity for +WASP-77Ab from JWST Thermal Emission Spectroscopy + August + Astrophysical Journal, +Letters + 2 + 953 + 10.3847/2041-8213/ace828 + 2023 + August, P. C., Bean, J. L., Zhang, +M., Lunine, J., Xue, Q., Line, M., & Smith, P. C. B. (2023). +Confirmation of Subsolar Metallicity for WASP-77Ab from JWST Thermal +Emission Spectroscopy. Astrophysical Journal, Letters, 953(2), L24. +https://doi.org/10.3847/2041-8213/ace828 + + + X-ray spectral modelling of the AGN obscuring +region in the CDFS: Bayesian model selection and +catalogue + Buchner + Astronomy & Astrophysics + 564 + 10.1051/0004-6361/201322971 + 2014 + Buchner, J., Georgakakis, A., Nandra, +K., Hsu, L., Rangel, C., Brightman, M., Merloni, A., Salvato, M., +Donley, J., & Kocevski, D. (2014). X-ray spectral modelling of the +AGN obscuring region in the CDFS: Bayesian model selection and +catalogue. Astronomy & Astrophysics, 564, A125. +https://doi.org/10.1051/0004-6361/201322971 + + + A systematic retrieval analysis of secondary +eclipse spectra. I. A comparison of atmospheric retrieval +techniques + Line + The Astrophysical Journal + 2 + 775 + 10.1088/0004-637X/775/2/137 + 2013 + Line, M. R., Wolf, A. S., Zhang, X., +Knutson, H., Kammer, J. A., Ellison, E., Deroo, P., Crisp, D., & +Yung, Y. L. (2013). A systematic retrieval analysis of secondary eclipse +spectra. I. A comparison of atmospheric retrieval techniques. The +Astrophysical Journal, 775(2), 137. +https://doi.org/10.1088/0004-637X/775/2/137 + + + Principal component analysis: A review and +recent developments + Jolliffe + Philosophical transactions of the royal +society A: Mathematical, Physical and Engineering +Sciences + 2065 + 374 + 10.1098/rsta.2015.0202 + 2016 + Jolliffe, I. T., & Cadima, J. +(2016). Principal component analysis: A review and recent developments. +Philosophical Transactions of the Royal Society A: Mathematical, +Physical and Engineering Sciences, 374(2065), 20150202. +https://doi.org/10.1098/rsta.2015.0202 + + + The HITRAN2016 molecular spectroscopic +database + Gordon + Journal of Quantitative Spectroscopy and +Radiative Transfer + 203 + 10.1016/j.jqsrt.2017.06.038 + 2017 + Gordon, I. E., Rothman, L. S., Hill, +C., Kochanov, R. V., Tan, Y., Bernath, P. F., Birk, M., Boudon, V., +Campargue, A., Chance, K., & others. (2017). The HITRAN2016 +molecular spectroscopic database. Journal of Quantitative Spectroscopy +and Radiative Transfer, 203, 3–69. +https://doi.org/10.1016/j.jqsrt.2017.06.038 + + + HITEMP, the high-temperature molecular +spectroscopic database + Rothman + Journal of Quantitative Spectroscopy and +Radiative Transfer + 15 + 111 + 10.1016/j.jqsrt.2010.05.001 + 2010 + Rothman, L. S., Gordon, I., Barber, +R., Dothe, H., Gamache, R. R., Goldman, A., Perevalov, V., Tashkun, S., +& Tennyson, J. (2010). HITEMP, the high-temperature molecular +spectroscopic database. Journal of Quantitative Spectroscopy and +Radiative Transfer, 111(15), 2139–2150. +https://doi.org/10.1016/j.jqsrt.2010.05.001 + + + Review of deep learning: Concepts, CNN +architectures, challenges, applications, future +directions + Alzubaidi + Journal of big Data + 8 + 10.1186/s40537-021-00444-8 + 2021 + Alzubaidi, L., Zhang, J., Humaidi, A. +J., Al-Dujaili, A., Duan, Y., Al-Shamma, O., Santamarı́a, J., Fadhel, M. +A., Al-Amidie, M., & Farhan, L. (2021). Review of deep learning: +Concepts, CNN architectures, challenges, applications, future +directions. Journal of Big Data, 8, 1–74. +https://doi.org/10.1186/s40537-021-00444-8 + + + JAX: Composable transformations of +Python+NumPy programs + Bradbury + 2018 + Bradbury, J., Frostig, R., Hawkins, +P., Johnson, M. J., Leary, C., Maclaurin, D., Necula, G., Paszke, A., +VanderPlas, J., Wanderman-Milne, S., & Zhang, Q. (2018). JAX: +Composable transformations of Python+NumPy programs (Version 0.3.13). +http://github.com/google/jax + + + 3D radiative transfer for exoplanet +atmospheres. gCMCRT: A GPU-accelerated MCRT code + Lee + The Astrophysical Journal + 2 + 929 + 10.3847/1538-4357/ac61d6 + 2022 + Lee, E. K., Wardenier, J. P., +Prinoth, B., Parmentier, V., Grimm, S. L., Baeyens, R., Carone, L., +Christie, D., Deitrick, R., Kitzmann, D., & others. (2022). 3D +radiative transfer for exoplanet atmospheres. gCMCRT: A GPU-accelerated +MCRT code. The Astrophysical Journal, 929(2), 180. +https://doi.org/10.3847/1538-4357/ac61d6 + + + Ooc_cuDNN: Accommodating convolutional neural +networks over GPU memory capacity + Ito + 2017 IEEE international conference on big +data (big data) + 10.1109/BigData.2017.8257926 + 2017 + Ito, Y., Matsumiya, R., & Endo, +T. (2017). Ooc_cuDNN: Accommodating convolutional neural networks over +GPU memory capacity. 2017 IEEE International Conference on Big Data (Big +Data), 183–192. +https://doi.org/10.1109/BigData.2017.8257926 + + + Corner.py: Scatterplot matrices in +python + Foreman-Mackey + The Journal of Open Source +Software + 2 + 1 + 10.21105/joss.00024 + 2016 + Foreman-Mackey, D. (2016). Corner.py: +Scatterplot matrices in python. The Journal of Open Source Software, +1(2), 24. https://doi.org/10.21105/joss.00024 + + + IGRINS at the discovery channel telescope and +gemini south + Mace + Ground-based and airborne instrumentation for +astronomy VII + 10702 + 10.1117/12.2312345 + 2018 + Mace, G., Sokal, K., Lee, J.-J., Oh, +H., Park, C., Lee, H., Good, J., MacQueen, P., Oh, J. S., Kaplan, K., +& others. (2018). IGRINS at the discovery channel telescope and +gemini south. Ground-Based and Airborne Instrumentation for Astronomy +VII, 10702, 204–221. +https://doi.org/10.1117/12.2312345 + + + ESPRESSO at VLT-on-sky performance and first +results + Pepe + Astronomy & Astrophysics + 645 + 10.1051/0004-6361/202038306 + 2021 + Pepe, F., Cristiani, S., Rebolo, R., +Santos, N., Dekker, H., Cabral, A., Di Marcantonio, P., Figueira, P., +Curto, G. L., Lovis, C., & others. (2021). ESPRESSO at VLT-on-sky +performance and first results. Astronomy & Astrophysics, 645, A96. +https://doi.org/10.1051/0004-6361/202038306 + + + On-sky commissioning of MAROON-x: A new +precision radial velocity spectrograph for gemini north + Seifahrt + Ground-based and airborne instrumentation for +astronomy VIII + 11447 + 10.1117/12.2561564 + 2020 + Seifahrt, A., Bean, J. L., Stürmer, +J., Kasper, D., Gers, L., Schwab, C., Zechmeister, M., Stefánsson, G., +Montet, B., Dos Santos, L. A., & others. (2020). On-sky +commissioning of MAROON-x: A new precision radial velocity spectrograph +for gemini north. Ground-Based and Airborne Instrumentation for +Astronomy VIII, 11447, 305–325. +https://doi.org/10.1117/12.2561564 + + + A temperature and abundance retrieval method +for exoplanet atmospheres + Madhusudhan + The Astrophysical Journal + 1 + 707 + 10.1088/0004-637X/707/1/24 + 2009 + Madhusudhan, N., & Seager, S. +(2009). A temperature and abundance retrieval method for exoplanet +atmospheres. The Astrophysical Journal, 707(1), 24. +https://doi.org/10.1088/0004-637X/707/1/24 + + + Modelling the effect of 3D temperature and +chemistry on the cross-correlation signal of transiting ultra-hot +Jupiters: a study of five chemical species on WASP-76b + Wardenier + Monthly Notices of the Royal Astronomical +Society + 4 + 525 + 10.1093/mnras/stad2586 + 0035-8711 + 2023 + Wardenier, J. P., Parmentier, V., +Line, M. R., & Lee, E. K. H. (2023). Modelling the effect of 3D +temperature and chemistry on the cross-correlation signal of transiting +ultra-hot Jupiters: a study of five chemical species on WASP-76b. +Monthly Notices of the Royal Astronomical Society, 525(4), 4942–4961. +https://doi.org/10.1093/mnras/stad2586 + + + Magnetic effects and 3D structure in +theoretical high-resolution transmission spectra of ultrahot jupiters: +The case of WASP-76b + Beltz + The Astronomical Journal + 6 + 165 + 10.3847/1538-3881/acd24d + 2023 + Beltz, H., Rauscher, E., Kempton, E. +M.-R., Malsky, I., & Savel, A. B. (2023). Magnetic effects and 3D +structure in theoretical high-resolution transmission spectra of +ultrahot jupiters: The case of WASP-76b. The Astronomical Journal, +165(6), 257. +https://doi.org/10.3847/1538-3881/acd24d + + + No umbrella needed: Confronting the +hypothesis of iron rain on WASP-76b with post-processed general +circulation models + Savel + The Astrophysical Journal + 1 + 926 + 10.3847/1538-4357/ac423f + 2022 + Savel, A. B., Kempton, E. M.-R., +Malik, M., Komacek, T. D., Bean, J. L., May, E. M., Stevenson, K. B., +Mansfield, M., & Rauscher, E. (2022). No umbrella needed: +Confronting the hypothesis of iron rain on WASP-76b with post-processed +general circulation models. The Astrophysical Journal, 926(1), 85. +https://doi.org/10.3847/1538-4357/ac423f + + + Time-resolved transmission spectroscopy of +the ultra-hot Jupiter WASP-189 b + Prinoth + Astronomy and Astrophysics + 678 + 10.1051/0004-6361/202347262 + 2023 + Prinoth, B., Hoeijmakers, H. J., +Pelletier, S., Kitzmann, D., Morris, B. M., Seifahrt, A., Kasper, D., +Korhonen, H. H., Burheim, M., Bean, J. L., Benneke, B., Borsato, N. W., +Brady, M., Grimm, S. L., Luque, R., Stürmer, J., & Thorsbro, B. +(2023). Time-resolved transmission spectroscopy of the ultra-hot Jupiter +WASP-189 b. Astronomy and Astrophysics, 678, A182. +https://doi.org/10.1051/0004-6361/202347262 + + + ExoMol molecular line lists–XXXV. A +rotation-vibration line list for hot ammonia + Coles + Monthly Notices of the Royal Astronomical +Society + 4 + 490 + 10.1093/mnras/stz2778 + 2019 + Coles, P. A., Yurchenko, S. N., & +Tennyson, J. (2019). ExoMol molecular line lists–XXXV. A +rotation-vibration line list for hot ammonia. Monthly Notices of the +Royal Astronomical Society, 490(4), 4638–4647. +https://doi.org/10.1093/mnras/stz2778 + + + Update of the HITRAN collision-induced +absorption section + Karman + Icarus + 328 + 10.1016/j.icarus.2019.02.034 + 2019 + Karman, T., Gordon, I. E., Der +Avoird, A. van, Baranov, Y. I., Boulet, C., Drouin, B. J., Groenenboom, +G. C., Gustafsson, M., Hartmann, J.-M., Kurucz, R. L., & others. +(2019). Update of the HITRAN collision-induced absorption section. +Icarus, 328, 160–175. +https://doi.org/10.1016/j.icarus.2019.02.034 + + + ExoMol line lists–III. An improved hot +rotation-vibration line list for HCN and HNC + Barber + Monthly Notices of the Royal Astronomical +Society + 2 + 437 + 10.1093/mnras/stt2011 + 2014 + Barber, R., Strange, J., Hill, C., +Polyansky, O., Mellau, G. C., Yurchenko, S., & Tennyson, J. (2014). +ExoMol line lists–III. An improved hot rotation-vibration line list for +HCN and HNC. Monthly Notices of the Royal Astronomical Society, 437(2), +1828–1835. https://doi.org/10.1093/mnras/stt2011 + + + ExoMol molecular line lists–XVI. The +rotation–vibration spectrum of hot H2S + Azzam + Monthly Notices of the Royal Astronomical +Society + 4 + 460 + 10.1093/mnras/stw1133 + 2016 + Azzam, A. A., Tennyson, J., +Yurchenko, S. N., & Naumenko, O. V. (2016). ExoMol molecular line +lists–XVI. The rotation–vibration spectrum of hot H2S. Monthly Notices +of the Royal Astronomical Society, 460(4), 4063–4074. +https://doi.org/10.1093/mnras/stw1133 + + + An accurate, extensive, and practical line +list of methane for the HITEMP database + Hargreaves + The Astrophysical Journal Supplement +Series + 2 + 247 + 10.3847/1538-4365/ab7a1a + 2020 + Hargreaves, R. J., Gordon, I. E., +Rey, M., Nikitin, A. V., Tyuterev, V. G., Kochanov, R. V., & +Rothman, L. S. (2020). An accurate, extensive, and practical line list +of methane for the HITEMP database. The Astrophysical Journal Supplement +Series, 247(2), 55. +https://doi.org/10.3847/1538-4365/ab7a1a + + + Retrieval survey of metals in six ultrahot +jupiters: Trends in chemistry, rain-out, ionization, and atmospheric +dynamics + Gandhi + The Astronomical Journal + 6 + 165 + 10.3847/1538-3881/accd65 + 2023 + Gandhi, S., Kesseli, A., Zhang, Y., +Louca, A., Snellen, I., Brogi, M., Miguel, Y., Casasayas-Barris, N., +Pelletier, S., Landman, R., & others. (2023). Retrieval survey of +metals in six ultrahot jupiters: Trends in chemistry, rain-out, +ionization, and atmospheric dynamics. The Astronomical Journal, 165(6), +242. https://doi.org/10.3847/1538-3881/accd65 + + + High-resolution atmospheric retrievals of +WASP-121b transmission spectroscopy with ESPRESSO: Consistent relative +abundance constraints across multiple epochs and +instruments + Maguire + Monthly Notices of the Royal Astronomical +Society + 1 + 519 + 10.1093/mnras/stac3388 + 2023 + Maguire, C., Gibson, N. P., Nugroho, +S. K., Ramkumar, S., Fortune, M., Merritt, S. R., & Mooij, E. de. +(2023). High-resolution atmospheric retrievals of WASP-121b transmission +spectroscopy with ESPRESSO: Consistent relative abundance constraints +across multiple epochs and instruments. Monthly Notices of the Royal +Astronomical Society, 519(1), 1030–1048. +https://doi.org/10.1093/mnras/stac3388 + + + Keras + Chollet + 2015 + Chollet, F. (2015). Keras. +https://keras.io. + + + ExoMol molecular line lists XXX: A complete +high-accuracy line list for water + Polyansky + Monthly Notices of the Royal Astronomical +Society + 2 + 480 + 10.1093/mnras/sty1877 + 2018 + Polyansky, O. L., Kyuberis, A. A., +Zobov, N. F., Tennyson, J., Yurchenko, S. N., & Lodi, L. (2018). +ExoMol molecular line lists XXX: A complete high-accuracy line list for +water. Monthly Notices of the Royal Astronomical Society, 480(2), +2597–2608. https://doi.org/10.1093/mnras/sty1877 + + + + + + diff --git a/joss.06104/10.21105.joss.06104.pdf b/joss.06104/10.21105.joss.06104.pdf new file mode 100644 index 0000000000..342d8ae633 Binary files /dev/null and b/joss.06104/10.21105.joss.06104.pdf differ diff --git a/joss.06104/paper.jats/10.21105.joss.06104.jats b/joss.06104/paper.jats/10.21105.joss.06104.jats new file mode 100644 index 0000000000..2293b61d03 --- /dev/null +++ b/joss.06104/paper.jats/10.21105.joss.06104.jats @@ -0,0 +1,1324 @@ + + +
+ + + + +Journal of Open Source Software +JOSS + +2475-9066 + +Open Journals + + + +6104 +10.21105/joss.06104 + +cortecs: A Python package for compressing +opacities + + + +https://orcid.org/0000-0002-2454-768X + +Savel +Arjun B. + + + +* + + +https://orcid.org/0000-0001-9907-7742 + +Bedell +Megan + + + + +https://orcid.org/0000-0002-1337-9051 + +Kempton +Eliza M.-R. + + + + + +Astronomy Department, University of Maryland, College Park, +4296 Stadium Dr., College Park, MD 207842 USA + + + + +Flatiron Institute, Simons Foundation, 162 Fifth Avenue, +New York, NY 10010, USA + + + + +* E-mail: + + +26 +8 +2023 + +9 +100 +6104 + +Authors of papers retain copyright and release the +work under a Creative Commons Attribution 4.0 International License (CC +BY 4.0) +2022 +The article authors + +Authors of papers retain copyright and release the work under +a Creative Commons Attribution 4.0 International License (CC BY +4.0) + + + +Python +astronomy +radiative transfer + + + + + + Summary +

The absorption and emission of light by exoplanet atmospheres + encode details of atmospheric composition, temperature, and dynamics. + Fundamentally, simulating these processes requires detailed knowledge + of the opacity of gases within an atmosphere. When modeling broad + wavelength ranges at high resolution, such opacity data, for even a + single gas, can take up multiple gigabytes of system random-access + memory (RAM). This aspect can be a limiting factor when considering + the number of gases to include in a simulation, the sampling strategy + used for inference, or even the architecture of the system used for + calculations. Here, we present cortecs, a + Python tool for compressing opacity data. + cortecs provides flexible methods for fitting + the temperature, pressure, and wavelength dependencies of opacity data + and for evaluating the opacity with accelerated, GPU-friendly methods. + The package is actively developed on GitHub + (https://github.com/arjunsavel/cortecs), + and it is available for download with pip and + conda.

+ + + Statement of need +

Observations with the latest high-resolution spectrographs (e.g., + IGRINS / Gemini South, ESPRESSO / VLT, MAROON-X / Gemini North; Mace + et al. + (2018); + Seifahrt et al. + (2020); + Pepe et al. + (2021)) + have motivated RAM-intensive simulations of exoplanet atmospheres at + high spectral resolution. cortecs enables these + simulations with more gases and on a broader range of computing + architectures by compressing opacity data.

+

Broadly, generating a spectrum to compare against recent + high-resolution data requires solving the radiative transfer equation + over tens of thousands of wavelength points (e.g., + Beltz + et al., 2023; + Gandhi + et al., 2023; + Line + et al., 2021; + Maguire + et al., 2023; + Prinoth + et al., 2023; + Savel + et al., 2022; + Wardenier + et al., 2023). To decrease computational runtime, some codes + have parallelized the problem on GPUs (e.g., + Lee + et al., 2022; + Line + et al., 2021). However, GPUs cannot in general hold large + amounts of data in their video random-access memory (VRAM) (e.g., + Ito + et al., 2017); only the cutting-edge, most expensive GPUs are + equipped with VRAM in excess of 30 GB (such as the NVIDIA A100 or + H100). RAM and VRAM management is therefore a clear concern when + producing high-resolution spectra.

+

How do we decrease the RAM footprint of these calculations? By far + the largest contributor to the RAM footprint, at least as measured on + disk, is the opacity data. For instance, the opacity data for a single + gas species across the wavelength range of the Immersion GRating + INfrared Spectrometer spectrograph (IGRINS, + Mace + et al., 2018) takes up 2.1 GB of non-volatile memory (i.e., the + file size is 2.1 GB) at float64 precision and + at a resolving power of 400,000 (as used in Line et al. + (2021); + with 39 temperature points and 18 pressure points, using, e.g., the + Polyansky et al. + (2018) + water opacity tables). In many cases, not all wavelengths need to be + loaded, e.g. if the user is down-sampling the resolution of their + opacity function. Even so, it stands to reason that decreasing the + amount of RAM/VRAM consumed by opacity data would strongly decrease + the total amount of RAM/VRAM consumed by the radiative transfer + calculation.

+

One solution is to isolate redundancy: While the wavelength + dependence of opacity is sharp for many gases, the temperature and + pressure dependencies are generally smooth and similar across + wavelengths (e.g., + Barber + et al., 2014; + Coles + et al., 2019; + Polyansky + et al., 2018). This feature implies that the opacity data + should be compressible without significant loss of accuracy at the + spectrum level.

+

While our benchmark case (see the Benchmark section below) + demonstrates the applicability of cortecs to + high-resolution opacity functions of molecular gases, the package is + general and the compression/decompression steps of the package can be + applied to any opacity data in HDF5 format that has pressure and + temperature dependence, such as the opacity of neutral atoms or ions. + Our benchmark only shows, however, that the amounts of error from our + compression technique is reasonable in the spectra of exoplanet + atmospheres at pressures greater than a microbar for a single + composition. This caveat is important to note for a few reasons:

+ + +

Based on error propagation, the error in the opacity function + may be magnified in the spectrum based on the number of cells that + are traced during radiative transfer. The number of spatial cells + used to simulate exoplanet atmospheres (in our case, 100) is small + enough that the cortecs error is not large + at the spectrum level.

+ + +

Exoplanet atmospheres are often modeled in hydrostatic + equilibrium at pressures greater than a microbar (e.g., + Barstow + et al., 2020; + Showman + et al., 2020). When modeling atmospheres in hydrostatic + equilibrium, the final spectrum essentially maps to the altitude + at which the gas becomes optically thick. If + cortecs-compressed opacities were used to + model an optically thin gas over large path lengths, however, then + smaller opacities would be more important. + cortecs tends to perform worse at modeling + opacity functions that jump from very low to very high opacities, + so it may not perform optimally in these optically thin + scenarios.

+ + +

The program may perform poorly for opacity functions with sharp + features in their temperature–pressure dependence (e.g., the Lyman + series transitions of hydrogen, + Kurucz, + 2017). That is, the data may require so many parameters to + be fit that the compression is no longer worthwhile.

+ + + + + Methods +

cortecs seeks to compress redundant + information by representing opacity data not as the opacity itself but + as fits to the opacity as a function of temperature and pressure. We + generally refer to this process as compression as + opposed to fitting to emphasize that we do not seek + to construct physically motivated, predictive, or comprehensible + models of the opacity function. Rather, we simply seek representations + of the opacity function that consume less RAM/VRAM. The compression + methods we use are lossy — the original opacity data + cannot be exactly recovered with our methods. We find that the loss of + accuracy is tolerable for at least the hot Jupiter emission + spectroscopy application (see Benchmark below).

+

We provide three methods of increasing complexity (and flexibility) + for compressing and decompressing opacity: polynomial-fitting, + principal components analysis (PCA, e.g., + Jolliffe + & Cadima, 2016) and neural networks (e.g., + Alzubaidi + et al., 2021). The default neural network architecture is a + fully connected neural network; the user can specify the desired + hyperparameters, such as number of layers, neurons per layer, and + activation function. Alternatively, any keras + model + (Chollet, + 2015) can be passed to the fitter. Each compression method is + paired with a decompression method for evaluating opacity as a + function of temperature, pressure, and wavelength. These decompression + methods are tailored for GPUs and are accelerated with the + JAX code transformation framework + (Bradbury + et al., 2018). An example of this reconstruction is shown in + [fig:example]. + In the figure, opacities less than + + 10−60 + are ignored. This is because, to become optically thick at a pressure + of 1 bar and temperature of 1000 K, a column would need to be nearly + + + 1034m + long. Here we show a brief derivation of this. The length of the + column, + + ds + is + + ds=τα, + where + + τ + is the optical depth, and + + α + is the absorption coefficient. Setting + + τ=1, + we have + + ds=1α. + The absorption coefficient is the product of the opacity and the + density of the gas: + + ds=1κλρ. + Therefore, + + ds=1κλρ. + The density of the gas + + ρ + is the pressure divided by the product of the temperature and the gas + constant: + + ρ=PkBTμ + for mean molecular weight + + μ. + This leads to the final equation for the column length: + + + ds=kBTμPκλ. + For CO, the mean molecular weight is 28.01 g/mol. Plugging in, we + arrive at + + ds≈1034m + (roughly + + 1017 + parsecs) for + + κλ=10−33 + , + which is equivalent to roughly a cross-section of + + + σλ=10−60 + .

+ +

Top panel: The original opacity function of CO + (Rothman + et al., 2010) (solid lines) and its + cortecs reconstruction (transparent lines) + over a large wavelength range and at multiple temperatures and + pressures. Bottom panel: the absolute residuals between the opacity + function and its cortecs reconstruction. + + + σλ + is the opacity, in units of square meters. We cut off the opacity at + + + 10−104, + explaining the shape of the residuals in teal and dark red. Note + that opacities less than + + 10−60 + are not generally relevant for the benchmark presented here; an + opacity of + + σλ=10−60 + would require a column nearly + + 1034m + long to become optically thick at a pressure of 1 bar and + temperature of 1000 K. +

+ + + + + Workflow +

A typical workflow with cortecs involves the + following steps:

+ + +

Acquiring: Download opacity data from a source such as the + ExoMol database + (Tennyson + et al., 2016) or the HITRAN database + (Gordon + et al., 2017).

+ + +

Fitting: Compress the opacity data with + cortecs’s fit + methods.

+ + +

Saving: Save the compressed opacity data (the fitted + parameters) to disk.

+ + +

Loading: Load the compressed opacity data from disk in whatever + program you’re applying the data—e.g., within your radiative + transfer code.

+ + +

Decompressing: Evaluate the opacity with + cortecs’s eval + methods.

+ + +

The accuracy of these fits may or may not be suitable for a given + application. It is important to test that the error incurred using + cortecs does not impact the results of your + application—for instance, by using the + cortecs.fit.metrics.calc_metrics function to + calculate the error incurred by the compression and by calculating + spectra with and without using + cortecs-compressed opacities. We provide an + example of such a benchmarking exercise below.

+ + + Benchmark: High-resolution retrieval of WASP-77Ab +

As a proof of concept, we perform a parameter inference exercise (a + “retrieval,” + Madhusudhan + & Seager, 2009) on the high-resolution thermal emission + spectrum of the fiducial hot Jupiter WASP-77Ab + (August + et al., 2023; + Line + et al., 2021; + Mansfield + et al., 2022) as observed at IGRINS. The retrieval pairs + PyMultiNest + (Buchner + et al., 2014) sampling with the CHIMERA + radiative transfer code + (Line + et al., 2013), with opacity from + + (Polyansky + et al., 2018), + + (Rothman + et al., 2010), + + (Hargreaves + et al., 2020), + + (Coles + et al., 2019), + + (Barber + et al., 2014), + + (Azzam + et al., 2016), and + + collision-induced absorption + (Karman + et al., 2019). The non-compressed retrieval uses the data and + retrieval framework from + (Line + et al., 2021), run in an upcoming article (Savel et al. 2024, + submitted). For this experiment, we use the PCA-based compression + scheme implemented in cortecs, preserving 2 + principal components and their corresponding weights as a function for + each wavelength as a lossy compression of the original opacity + data.

+

Using cortecs, we compress the input opacity + files by a factor of 13. These opacity data (as described earlier in + the paper) were originally stored as 2.1 GB .h5 files containing 39 + temperature points, 18 pressure points, and 373,260 wavelength points. + The compressed opacity data are stored as a 143.1 MB .npz file, + including the PCA coefficients and PCA vectors (which are reused for + each wavelength point). These on-disk memory quotes are relatively + faithful to the in-memory RAM footprint of the data when stored as + numpy arrays (2.1 GB for the uncompressed data + and 160 MB for the compressed data). Reading in the original files + takes 1.1 + + ± + 0.1 seconds, while reading in the compressed files takes 24.4 + + + ± + 0.3 ms. Accessing/evaluating a single opacity value takes 174.0 + + + ± + 0.5 ns for the uncompressed data and 789 + + + ± + 5 ns for the compressed data. All of these timing experiments are + performed on a 2021 MacBook Pro with an Apple M1 Pro chip and 16 GB of + RAM.

+

Importantly, we find that our compressed-opacity retrieval yields + posterior distributions (as plotted by the + corner package, + Foreman-Mackey, + 2016) and Bayesian evidences that are consistent with those + from the benchmark retrieval using uncompressed opacity + ([fig:corner]) + within a comparable runtime. The two posterior distributions exhibit + slightly different substructure, which we attribute to the compressed + results requiring 10% more samples to converge (about 5 hours of extra + runtime on a roughly 2 day-long calculation) and residual differences + between the compressed and uncompressed opacities. The results from + this exercise indicate that our compression/decompression scheme is + accurate enough to be used in at least some high-resolution + retrievals.

+ +

The posterior distributions for our baseline WASP-77Ab + retrieval (teal) and our retrieval using opacities compressed by + cortecs (gold). +

+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
MethodCompression factorMedian absolute deviationCompression time (s)Decompression time (s)
PCA130.302.6 + + ×1012.3 + + ×102
Polynomials440.247.8 + + ×1023.6 + + ×103
Neural network92.61.4 + + ×1073.6 + + ×104
+ +

Comparison of compression methods used for the full HITEMP CO line + list + (Rothman + et al., 2010) over the IGRINS wavelength range at a resolving + power of 250,000, cumulative for all data points. Note that the neural + network compression performance and timings are only assessed at a + single wavelength point and extrapolated over the full wavelength + range.

+ + + Acknowledgements +

A.B.S. and E.M-R.K. acknowledge support from the Heising-Simons + Foundation. We thank Max Isi for helpful discussions.

+ + + + + + + + + KuruczRobert L + + Including all the lines: Data releases for spectra and opacities + Canadian Journal of Physics + NRC Research Press + 2017 + 95 + 9 + 10.1139/cjp-2016-0794 + 825 + 827 + + + + + + ShowmanAdam P + TanXianyu + ParmentierVivien + + Atmospheric dynamics of hot giant planets and brown dwarfs + Space Science Reviews + Springer + 2020 + 216 + 10.1007/s11214-020-00758-8 + 1 + 83 + + + + + + BarstowJoanna K + ChangeatQuentin + GarlandRyan + LineMichael R + RocchettoMarco + WaldmannIngo P + + A comparison of exoplanet spectroscopic retrieval tools + Monthly Notices of the Royal Astronomical Society + Oxford University Press + 2020 + 493 + 4 + 10.1093/mnras/staa548 + 4884 + 4909 + + + + + + LineMichael R + BrogiMatteo + BeanJacob L + GandhiSiddharth + ZaleskyJoseph + ParmentierVivien + SmithPeter + MaceGregory N + MansfieldMegan + KemptonEliza M-R + others + + A solar c/o and sub-solar metallicity in a hot jupiter atmosphere + Nature + Nature Publishing Group UK London + 2021 + 598 + 7882 + 10.1038/s41586-021-03912-6 + 580 + 584 + + + + + + TennysonJonathan + YurchenkoSergei N + Al-RefaieAhmed F + BartonEmma J + ChubbKaty L + ColesPhillip A + DiamantopoulouS + GormanMaire N + HillChristian + LamAden Z + others + + The ExoMol database: Molecular line lists for exoplanet and other hot atmospheres + Journal of Molecular Spectroscopy + Elsevier + 2016 + 327 + 10.1016/j.jms.2016.05.002 + 73 + 94 + + + + + + MansfieldMegan + WiserLindsey + StevensonKevin B + SmithPeter + LineMichael R + BeanJacob L + FortneyJonathan J + ParmentierVivien + KemptonEliza M-R + ArcangeliJacob + others + + Confirmation of water absorption in the thermal emission spectrum of the hot jupiter WASP-77Ab with HST/WFC3 + The Astronomical Journal + IOP Publishing + 2022 + 163 + 6 + 10.3847/1538-3881/ac658f + 261 + + + + + + + AugustPrune C. + BeanJacob L. + ZhangMichael + LunineJonathan + XueQiao + LineMichael + SmithPeter C. B. + + Confirmation of Subsolar Metallicity for WASP-77Ab from JWST Thermal Emission Spectroscopy + Astrophysical Journal, Letters + 202308 + 953 + 2 + https://arxiv.org/abs/2305.07753 + 10.3847/2041-8213/ace828 + L24 + + + + + + + BuchnerJ + GeorgakakisA + NandraK + HsuL + RangelC + BrightmanM + MerloniA + SalvatoM + DonleyJ + KocevskiD + + X-ray spectral modelling of the AGN obscuring region in the CDFS: Bayesian model selection and catalogue + Astronomy & Astrophysics + EDP Sciences + 2014 + 564 + 10.1051/0004-6361/201322971 + A125 + + + + + + + LineMichael R + WolfAaron S + ZhangXi + KnutsonHeather + KammerJoshua A + EllisonElias + DerooPieter + CrispDave + YungYuk L + + A systematic retrieval analysis of secondary eclipse spectra. I. A comparison of atmospheric retrieval techniques + The Astrophysical Journal + IOP Publishing + 2013 + 775 + 2 + 10.1088/0004-637X/775/2/137 + 137 + + + + + + + JolliffeIan T + CadimaJorge + + Principal component analysis: A review and recent developments + Philosophical transactions of the royal society A: Mathematical, Physical and Engineering Sciences + The Royal Society Publishing + 2016 + 374 + 2065 + 10.1098/rsta.2015.0202 + 20150202 + + + + + + + GordonIouli E + RothmanLaurence S + HillChristian + KochanovRoman V + TanY + BernathPeter F + BirkManfred + BoudonV + CampargueAlain + ChanceKV + others + + The HITRAN2016 molecular spectroscopic database + Journal of Quantitative Spectroscopy and Radiative Transfer + Elsevier + 2017 + 203 + 10.1016/j.jqsrt.2017.06.038 + 3 + 69 + + + + + + RothmanLaurence S + GordonIE + BarberRJ + DotheH + GamacheRobert R + GoldmanA + PerevalovVI + TashkunSA + TennysonJ + + HITEMP, the high-temperature molecular spectroscopic database + Journal of Quantitative Spectroscopy and Radiative Transfer + Elsevier + 2010 + 111 + 15 + 10.1016/j.jqsrt.2010.05.001 + 2139 + 2150 + + + + + + AlzubaidiLaith + ZhangJinglan + HumaidiAmjad J + Al-DujailiAyad + DuanYe + Al-ShammaOmran + Santamarı́aJosé + FadhelMohammed A + Al-AmidieMuthana + FarhanLaith + + Review of deep learning: Concepts, CNN architectures, challenges, applications, future directions + Journal of big Data + Springer + 2021 + 8 + 10.1186/s40537-021-00444-8 + 1 + 74 + + + + + + BradburyJames + FrostigRoy + HawkinsPeter + JohnsonMatthew James + LearyChris + MaclaurinDougal + NeculaGeorge + PaszkeAdam + VanderPlasJake + Wanderman-MilneSkye + ZhangQiao + + JAX: Composable transformations of Python+NumPy programs + 2018 + http://github.com/google/jax + + + + + + LeeElspeth KH + WardenierJoost P + PrinothBibiana + ParmentierVivien + GrimmSimon L + BaeyensRobin + CaroneLudmila + ChristieDuncan + DeitrickRussell + KitzmannDaniel + others + + 3D radiative transfer for exoplanet atmospheres. gCMCRT: A GPU-accelerated MCRT code + The Astrophysical Journal + IOP Publishing + 2022 + 929 + 2 + 10.3847/1538-4357/ac61d6 + 180 + + + + + + + ItoYuki + MatsumiyaRyo + EndoToshio + + Ooc_cuDNN: Accommodating convolutional neural networks over GPU memory capacity + 2017 IEEE international conference on big data (big data) + IEEE + 2017 + 10.1109/BigData.2017.8257926 + 183 + 192 + + + + + + Foreman-MackeyDaniel + + Corner.py: Scatterplot matrices in python + The Journal of Open Source Software + The Open Journal + 201606 + 1 + 2 + https://doi.org/10.21105/joss.00024 + 10.21105/joss.00024 + 24 + + + + + + + MaceGregory + SokalKimberly + LeeJae-Joon + OhHeeyoung + ParkChan + LeeHanshin + GoodJohn + MacQueenPhillip + OhJae Sok + KaplanKyle + others + + IGRINS at the discovery channel telescope and gemini south + Ground-based and airborne instrumentation for astronomy VII + SPIE + 2018 + 10702 + 10.1117/12.2312345 + 204 + 221 + + + + + + PepeF + CristianiStefano + ReboloR + SantosNC + DekkerH + CabralA + Di MarcantonioPaolo + FigueiraP + CurtoG Lo + LovisC + others + + ESPRESSO at VLT-on-sky performance and first results + Astronomy & Astrophysics + EDP Sciences + 2021 + 645 + 10.1051/0004-6361/202038306 + A96 + + + + + + + SeifahrtAndreas + BeanJacob L + StürmerJulian + KasperDavid + GersLuke + SchwabChristian + ZechmeisterMathias + StefánssonGumundur + MontetBen + Dos SantosLeonardo A + others + + On-sky commissioning of MAROON-x: A new precision radial velocity spectrograph for gemini north + Ground-based and airborne instrumentation for astronomy VIII + SPIE + 2020 + 11447 + 10.1117/12.2561564 + 305 + 325 + + + + + + MadhusudhanN + SeagerSara + + A temperature and abundance retrieval method for exoplanet atmospheres + The Astrophysical Journal + IOP Publishing + 2009 + 707 + 1 + 10.1088/0004-637X/707/1/24 + 24 + + + + + + + WardenierJoost P + ParmentierVivien + LineMichael R + LeeElspeth K H + + Modelling the effect of 3D temperature and chemistry on the cross-correlation signal of transiting ultra-hot Jupiters: a study of five chemical species on WASP-76b + Monthly Notices of the Royal Astronomical Society + 202308 + 525 + 4 + 0035-8711 + https://doi.org/10.1093/mnras/stad2586 + 10.1093/mnras/stad2586 + 4942 + 4961 + + + + + + BeltzHayley + RauscherEmily + KemptonEliza M-R + MalskyIsaac + SavelArjun B + + Magnetic effects and 3D structure in theoretical high-resolution transmission spectra of ultrahot jupiters: The case of WASP-76b + The Astronomical Journal + IOP Publishing + 2023 + 165 + 6 + 10.3847/1538-3881/acd24d + 257 + + + + + + + SavelArjun B + KemptonEliza M-R + MalikMatej + KomacekThaddeus D + BeanJacob L + MayErin M + StevensonKevin B + MansfieldMegan + RauscherEmily + + No umbrella needed: Confronting the hypothesis of iron rain on WASP-76b with post-processed general circulation models + The Astrophysical Journal + IOP Publishing + 2022 + 926 + 1 + 10.3847/1538-4357/ac423f + 85 + + + + + + + PrinothB. + HoeijmakersH. J. + PelletierS. + KitzmannD. + MorrisB. M. + SeifahrtA. + KasperD. + KorhonenH. H. + BurheimM. + BeanJ. L. + BennekeB. + BorsatoN. W. + BradyM. + GrimmS. L. + LuqueR. + StürmerJ. + ThorsbroB. + + Time-resolved transmission spectroscopy of the ultra-hot Jupiter WASP-189 b + Astronomy and Astrophysics + 202310 + 678 + https://arxiv.org/abs/2308.04523 + 10.1051/0004-6361/202347262 + A182 + + + + + + + ColesPhillip A + YurchenkoSergei N + TennysonJonathan + + ExoMol molecular line lists–XXXV. A rotation-vibration line list for hot ammonia + Monthly Notices of the Royal Astronomical Society + Oxford University Press + 2019 + 490 + 4 + 10.1093/mnras/stz2778 + 4638 + 4647 + + + + + + KarmanTijs + GordonIouli E + Der AvoirdAd van + BaranovYury I + BouletChristian + DrouinBrian J + GroenenboomGerrit C + GustafssonMagnus + HartmannJean-Michel + KuruczRobert L + others + + Update of the HITRAN collision-induced absorption section + Icarus + Elsevier + 2019 + 328 + 10.1016/j.icarus.2019.02.034 + 160 + 175 + + + + + + BarberRJ + StrangeJK + HillC + PolyanskyOL + MellauG Ch + YurchenkoSN + TennysonJonathan + + ExoMol line lists–III. An improved hot rotation-vibration line list for HCN and HNC + Monthly Notices of the Royal Astronomical Society + Oxford University Press + 2014 + 437 + 2 + 10.1093/mnras/stt2011 + 1828 + 1835 + + + + + + AzzamAla’a AA + TennysonJonathan + YurchenkoSergei N + NaumenkoOlga V + + ExoMol molecular line lists–XVI. The rotation–vibration spectrum of hot H2S + Monthly Notices of the Royal Astronomical Society + Oxford University Press + 2016 + 460 + 4 + 10.1093/mnras/stw1133 + 4063 + 4074 + + + + + + HargreavesRobert J + GordonIouli E + ReyMichael + NikitinAndrei V + TyuterevVladimir G + KochanovRoman V + RothmanLaurence S + + An accurate, extensive, and practical line list of methane for the HITEMP database + The Astrophysical Journal Supplement Series + IOP Publishing + 2020 + 247 + 2 + 10.3847/1538-4365/ab7a1a + 55 + + + + + + + GandhiSiddharth + KesseliAurora + ZhangYapeng + LoucaAmy + SnellenIgnas + BrogiMatteo + MiguelYamila + Casasayas-BarrisNúria + PelletierStefan + LandmanRico + others + + Retrieval survey of metals in six ultrahot jupiters: Trends in chemistry, rain-out, ionization, and atmospheric dynamics + The Astronomical Journal + IOP Publishing + 2023 + 165 + 6 + 10.3847/1538-3881/accd65 + 242 + + + + + + + MaguireCathal + GibsonNeale P + NugrohoStevanus K + RamkumarSwaetha + FortuneMark + MerrittStephanie R + MooijErnst de + + High-resolution atmospheric retrievals of WASP-121b transmission spectroscopy with ESPRESSO: Consistent relative abundance constraints across multiple epochs and instruments + Monthly Notices of the Royal Astronomical Society + Oxford University Press + 2023 + 519 + 1 + 10.1093/mnras/stac3388 + 1030 + 1048 + + + + + + CholletFrançois + + Keras + https://keras.io + 2015 + + + + + + PolyanskyOleg L + KyuberisAleksandra A + ZobovNikolai F + TennysonJonathan + YurchenkoSergei N + LodiLorenzo + + ExoMol molecular line lists XXX: A complete high-accuracy line list for water + Monthly Notices of the Royal Astronomical Society + Oxford University Press + 2018 + 480 + 2 + 10.1093/mnras/sty1877 + 2597 + 2608 + + + + +
diff --git a/joss.06104/paper.jats/example_application.png b/joss.06104/paper.jats/example_application.png new file mode 100644 index 0000000000..922b31cfbd Binary files /dev/null and b/joss.06104/paper.jats/example_application.png differ diff --git a/joss.06104/paper.jats/pca_compress.png b/joss.06104/paper.jats/pca_compress.png new file mode 100644 index 0000000000..aac3dfd429 Binary files /dev/null and b/joss.06104/paper.jats/pca_compress.png differ