Pith. sign in

REVIEW 4 major objections 6 minor 26 references

Icarus 3.0: Dynamic Heliosphere Modelling

T0 review · 4 major / 6 minor · reviewed 2026-08-10 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read Icarus 3.0 shows that refreshing the solar-wind boundary every two hours produces heliospheric forecasts that match in-situ spacecraft data better than a fixed steady wind.

desk verdict Solid engineering upgrade for Icarus with a plausible but visually supported only claim that dynamic driving beats steady driving for space weather. read the letter →

arxiv 2501.15923 v1 pith:56MNDRWL submitted 2025-01-27 astro-ph.SR physics.comp-phphysics.plasm-phphysics.space-ph

Tinatin Baratashvili , Beatrice Popescu Braileanu , Fabio Bacchini , Rony Keppens , Stefaan Poedts This is my paper · ORCID
classification astro-ph.SRphysics.comp-phphysics.plasm-phphysics.space-ph
keywords spaceweathersolarwindinnerheliospherecoronalmassejectionsmagnetohydrodynamicstime-dependentboundaryconditionsco-rotatinginteractionregionsIcarusmodel
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper argues that a heliospheric space-weather model driven by a time-varying inner boundary at 0.1 AU produces a more realistic solar wind than the standard practice of freezing one magnetogram and running a steady wind for days or weeks. The authors upgrade the Icarus model so the coronal boundary is recomputed on successive magnetograms, stacked in time, and linearly interpolated during the magnetohydrodynamic simulation. In two validation periods—the September-October 2019 solar wind and the June-July 2015 solar wind with five coronal mass ejections—the dynamic runs match spacecraft velocity, density, and magnetic-field time series more closely than steady runs, particularly catching high-speed streams that the steady runs miss and staying accurate after the boundary magnetogram would have gone stale. The authors conclude that dynamic driving makes Icarus better suited for space weather forecasting at no significant extra computational cost.

What carries the argument

The key mechanism is the time-dependent inner boundary condition at $r = 0.1$ AU. An automated pipeline computes the coronal model for each selected magnetogram, converts each 2D boundary shell (radial velocity, number density, pressure, and radial magnetic field) into a data file, and stacks the shells along a time axis; the MHD solver then reads this lookup table and linearly interpolates between consecutive snapshots at every simulation time step. A new boundary-condition type, bc_icarus, extends data-driven boundary driving so that cone-model CMEs can be superposed on the varying solar wind, and the particle-sampling module was adapted to stretched spherical adaptive grids so spacecraft positions can be sampled through the evolving flow.

What would settle it

Take a third, independent interval with multiple high-speed streams and in-situ data from at least two spacecraft, run the same Icarus 3.0 setup with a 2-hour refresh cadence and with the best single steady magnetogram, and compare both runs to observations with a quantitative score such as correlation or root-mean-square error over the full 30-40 day window. The paper's central claim fails if the dynamic run is not consistently closer to the observed velocity, density, and magnetic-field time series, especially more than a week after the steady magnetogram's date.

Watch

Extended reading notes

Core claim

The central claim is that switching the inner boundary of the heliospheric model from a fixed to a time-evolving prescription improves agreement with in-situ observations. Icarus 3.0 repeatedly runs an empirical coronal model on successive magnetograms, stacks the resulting 0.1 AU boundary states into a time series, and linearly interpolates between them during the simulation. In the 2019 case the dynamic wind at Earth and at a second spacecraft reproduces high-speed streams and density variations that the steady wind misses; in the 2015 case, CMEs propagated through the dynamic background yield time-series signatures closer to observed data once the wind is restored. The authors state that the dynamic solar wind agrees better with observations than steady driving and therefore that the upgraded Icarus is better suited for space weather forecasting.

Load-bearing premise

The load-bearing premise is that the empirical coronal model, with parameters tuned on one period and fed by the available magnetograms, produces a boundary at 0.1 AU that faithfully represents the real evolving solar wind; if that boundary mapping is biased, dynamic driving simply propagates the bias rather than improving the forecast.

Editorial extensions

If this is right

  • Long-horizon solar wind simulations stay accurate for weeks rather than degrading as the input magnetogram ages.
  • High-speed streams that a fixed-magnetogram run misses, such as the early-October 2019 stream at Earth, appear in the dynamic run and improve the number-density profile.
  • CMEs injected into a dynamic background produce Earth time-series signatures closer to observations than the same CMEs in a steady wind, and the improvement persists after the ejecta have passed.
  • The computational cost of dynamic driving is comparable to steady driving in the two test cases, while adding adaptive mesh refinement costs extra time but sharpens CME and shock structure.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A direct, testable extension is to vary the boundary-update cadence (e.g. 1 hour versus 1 day) and score the forecasts quantitatively; the paper compares visual profiles at two fixed cadences, so the optimal refresh rate is not established.
  • Because dynamic driving prevents the boundary from going stale, the same machinery should extend the usable heliospheric domain beyond 2 AU toward Jupiter, a step the paper names as future work; the added cost is then dominated by the repeated coronal-model runs rather than by the heliospheric solver.
  • If the improvement holds for active periods, operational forecast chains could in principle update their ambient wind every few hours at negligible extra heliospheric runtime, with magnetogram delivery and coronal-model processing becoming the practical bottleneck.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. The paper presents Icarus 3.0, an upgrade of the heliospheric MHD model that replaces steady inner boundary conditions at 0.1 AU with time-dependent boundary driving derived from repeated WSA coronal runs with evolving magnetograms. The authors implement a new bc_icarus boundary condition, port the code to MPI-AMRVAC 3.0, and extend particle sampling to stretched spherical grids. The model is tested against two periods: September-October 2019 (solar wind only, compared with OMNI and STEREO-A data) and June-July 2015 (solar wind with five consecutive CMEs, compared with OMNI data). The authors conclude that dynamic boundary driving yields a more accurate solar wind and that Icarus is better suited for space weather forecasting than a steady-wind model.

Significance. If established, the central claim would be a valuable step toward long-horizon heliospheric forecasting because it removes the known limitation of stale magnetogram boundary conditions and enables continuous updating of the inner boundary. The engineering is a strength: the code is publicly maintained, the old and new boundary treatments are explicitly compared, and the use of built-in MPI-AMRVAC 3.0 functionality is a good practice. However, the paper's central comparison is visual and in-sample, and both test cases involve tuning on the validation interval. The significance of the claimed improvement is therefore not yet demonstrated; quantitative skill scores and an out-of-sample test would substantially strengthen the manuscript.

major comments (4)
  1. [3.1; Conclusions] The central claim that dynamic boundary driving yields a more accurate solar wind is supported only by visual comparison. Section 3.1 states "The results are compared visually instead of running the various models due to noise in the observed data," and Figures 4 and 5 are presented without any skill scores (e.g., RMSE, correlation, or stream-arrival phase error). The abstract and conclusions convert this into the unqualified statements that dynamic results "are more accurate" and that Icarus is "better suited for space weather forecasting." Please add quantitative error metrics for the velocity, density, and magnetic field time series over defined evaluation windows, and state explicitly which intervals are used for the comparison. Without such metrics, the headline conclusion is not established.
  2. [3.1, 3.2] The validation is in-sample. For the 2019 case, the paper notes that "the coronal model was modified to account for this high-speed stream in Wijsen et al. (2021), and the same modified WSA parameters are used to generate all boundary files"; thus the dynamic simulation benefits from parameters tuned on the very interval used for evaluation. For the 2015 case, the GONGz magnetogram product is selected because it "better agrees with the observed data" on that same interval (Figure 6). The observed improvement of dynamic over steady driving could therefore reflect tuning to the validation data rather than a general property of dynamic boundary driving. A demonstration on an out-of-sample interval, or at least a clear discussion of this limitation, is required before claiming general superiority.
  3. [3.2, Fig. 8] The most striking dynamic-over-steady improvement in the 2015 case occurs after 1 July, which the paper itself attributes to the steady magnetogram being "majorly outdated" (two weeks old). This comparison conflates boundary-data freshness with the dynamic-driving method: a steady run using a more recent magnetogram for the late interval, or an evaluation restricted to the period where the steady boundary is still representative, would be needed to isolate the effect of time-dependent updating. As it stands, the statement that "dynamic simulations show improvement compared to the steady ones" is not an apples-to-apples comparison over the full time series.
  4. [Abstract; 3.2, Fig. 8] The claim that CMEs propagating through the dynamic solar wind "produce more similar signatures in the time-series data than in the steady solar wind" is likewise unsupported quantitatively. Only one multi-CME event is shown, the comparison is visual, and the magnetic field signatures are acknowledged to be underestimated because the cone CME model is non-magnetized. Please provide CME arrival-time and speed errors (e.g., against the DONKI/OMNI catalog) for the dynamic and steady runs, and state the number of events used.
minor comments (6)
  1. [Throughout] Several instances of "di fferent" with a non-breaking space appear in the text (e.g., Section 1, "di fferent sources"); please replace with "different."
  2. [3.1] The sentence "The results are compared visually instead of running the various models due to noise in the observed data" is confusing because the models have already been run; please rephrase to "instead of computing quantitative metrics due to noise in the observed data."
  3. [3.2] The cadence of the WSA boundary updates for the 2015 case is not stated; please specify it, as is done for the 2019 case (two hours).
  4. [4] The statement that dynamic simulations were "slightly faster" than steady ones is based on single runs; please add a note that these are single wall-clock measurements and are not statistically robust.
  5. [Figures 4, 5, 6, 8] The captions are inconsistent about the time averaging of the observed OMNI data; please state for each figure whether the plotted black curves are 1-minute data and, if so, what filtering or decimation was applied.
  6. [Table 2] The "Time at 0.1 AU" column uses date-time values without an explicit time zone; please state in the caption that the times are UT, as assumed from the DONKI catalog.

Circularity Check

2 steps flagged · score 4.0 of 10

Validation evidence for the dynamic-over-steady claim is partly in-sample (WSA parameters tuned to the 2019 interval; GONGz selected on the 2015 interval), but the dynamic-vs-steady comparison itself is not forced by those choices.

  1. fitted input called prediction [Section 3.1, September–October 2019 comparison (paragraph after Fig. 4)]
    "as mentioned earlier, the coronal model was modified to account for this high-speed stream in Wijsen et al. (2021), and the same modified WSA parameters are used to generate all boundary files with the WSA model."

    The WSA coronal parameters were adjusted in prior work specifically to reproduce the high-speed stream in the September–October 2019 interval that is later used as the validation target. The dynamic boundary files are generated with those same tuned parameters, so the dynamic run's better reproduction of the observed speed and density profiles is not an independent forecast of that interval: it is the heliospheric propagation of a boundary input that was already fitted to the observed target. The steady run uses the same tuned parameters, so the dynamic-versus-steady comparison is not vacuous, but the absolute agreement used to support the paper's forecasting claim is in-sample and partly determined by the fit.

  2. other [Section 3.2, June–July 2015 GONG product choice (text after Fig. 6)]
    "the modelled data with the zero-point corrected magnetogram better agrees with the observed data than the standard case. ... Therefore, to assess the results of our simulations, we chose the GONGz product."

    The magnetogram product for the 2015 validation interval was selected because it gives the best agreement with the same OMNI observations against which the dynamic simulation is later judged. The subsequent dynamic-versus-steady comparison in Fig. 8 shares the selected GONGz input, so the central comparison is not forced by the choice; however, the absolute validation that GONGz is better and that the dynamic wind is more accurate is performed on the same interval used to make the selection, removing an independent out-of-sample confirmation of the claim.

full rationale

This paper contains no analytical derivation whose conclusion is identical to its premise; the Icarus 3.0 upgrade is a numerical implementation, and the dynamic-versus-steady comparison is a genuine simulation experiment. The main circularity-adjacent issue is in the validation design. For the 2019 case, the WSA coronal parameters were modified in Wijsen et al. (2021) specifically to reproduce the high-speed stream in that interval, and the same parameters generate all dynamic boundary files, so the agreement of the dynamic run with OMNI and Stereo-A data is partly a propagation of a tuned input rather than an independent forecast. For the 2015 case, the GONGz magnetogram product was chosen because it best matches the same OMNI observations later used to judge the simulations. Neither choice makes the dynamic-over-steady conclusion true by construction, because both steady and dynamic simulations share the same tuned boundary model and the same magnetogram product; the dynamic advantage could still arise from genuinely updated boundary data. However, the paper's broader claim that Icarus is 'better suited for space weather forecasting' is supported only by visual, in-sample comparisons (Section 3.1 explicitly states 'The results are compared visually instead of running the various models due to noise in the observed data'), with no out-of-sample forecast or quantitative skill score. This is a validation-evidence weakness and a partial circularity of the input-tuning kind, but not a case where the derivation is equivalent to its inputs; hence a moderate score of 4.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The central claim rests on the empirical WSA coronal mapping, on model parameters (gamma, relaxation time, WSA tuning, GONG product choice) that are taken from prior work or selected with knowledge of the validation data, and on the assumption that linear interpolation between boundary snapshots preserves accuracy. No new physical entities are introduced.

free parameters (2)
  • WSA coronal model parameters (modified per Wijsen et al. 2021) = not stated
    Tuned previously to reproduce the September 2019 high-speed stream, then used to generate all 454 boundary files; the validation is therefore not blind.
  • Relaxation time = 8 days
    User-set spin-up chosen so slow wind reaches the outer boundary; affects the initial transient in both steady and dynamic runs.
assumptions (4)
  • domain assumption Ideal MHD equations with gamma = 1.5 and co-rotating frame source terms adequately capture solar wind acceleration in the inner heliosphere.
    Used throughout Section 2; the reduced gamma is a known simple proxy for heating adopted from Pomoell & Poedts (2018).
  • domain assumption WSA output at 0.1 AU, with linear interpolation between magnetogram snapshots, is a sufficient representation of the real evolving boundary.
    Section 3 procedure; if the empirical WSA mapping is biased, dynamic driving inherits the bias.
  • ad hoc to paper An 8-day relaxation produces a representative initial heliosphere for each case.
    Section 3, step 5; fixed by the authors and not tested for sensitivity.
  • domain assumption The non-magnetized cone CME model is sufficient for testing solar wind background effects on CME signatures.
    Section 3.2; the authors note the magnetic field is underestimated because the CME is not magnetized.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Icarus 3.0: Dynamic Heliosphere Modelling." pith.science (2026). https://pith.science/paper/56MNDRWL

@misc{pith2026250115923,
  author       = {Pith},
  title        = {Pith review of: Icarus 3.0: Dynamic Heliosphere Modelling},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/56MNDRWL}},
  note         = {Machine review of arXiv:2501.15923}
}
read the original abstract

Space weather predictions are necessary to avoid damage caused by intense geomagnetic storms. Such strong storms are usually caused by a co-rotating interaction region (CIR) passing at Earth or by the arrival of strong coronal mass ejections (CMEs). To mitigate the damage, the effect of propagating CMEs in the solar wind must be estimated accurately at Earth and other locations. Modelling solar wind accurately is crucial for space weather predictions, as it is the medium for CME propagation. The Icarus heliospheric modelling tool is upgraded to handle dynamic inner heliospheric driving instead of using steady boundary conditions. The ideal magnetohydrodynamic (MHD) solver and the automated grid-adaptivity are adjusted to the latest MPI-AMRVAC version. The inner boundary conditions, prescribed at 0.1 AU for the heliospheric model, are updated time-dependently throughout the simulation. The coronal model is computed repeatedly for selected magnetograms. The particle sampling within MPI-AMRVAC is extended to handle stretched spherical grid information. The solar wind obtained in the simulation is dynamic and shows significant variations throughout the evolution. When comparing the results with the observations, the dynamic solar wind results are more accurate than previous results obtained with purely steady boundary driving. The CMEs propagated through the dynamic solar wind background produce more similar signatures in the time-series data than in the steady solar wind. Dynamic boundary driving in Icarus results in a more self-consistent solar wind evolution in the inner heliosphere. The obtained space weather modelling tool for dynamic solar wind and CME simulations is better suited for space weather forecasting than a steady solar wind model.

Figures

Figures reproduced from arXiv: 2501.15923 by the authors.

Figure 2
Figure 2. Radial velocity (upper panel) and number density (lower [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. The equatorial planes from the same Icarus simulation at 15, 20, 25, 30, 35, and 40 days in the simulation from the upper [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figure 5
Figure 5. Radial velocity (upper panel) and number density (lower [PITH_FULL_IMAGE:figures/full_fig_p005_5.png] view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Radial velocity (upper panel) and number density (lower [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 6
Figure 6. Figure 6: Medium resolution simulation results performed with [PITH_FULL_IMAGE:figures/full_fig_p006_6.png]
Figure 7
Figure 7. Figure 7: The equatorial planes from the same Icarus simulations at 9, 10.3, and 11.5 days in the simulation. The upper row shows the CME density and ∇ · V values in the code units overplotted with the refined mesh. The bottom row shows the radial velocities for the same snapsho…
Figure 8
Figure 8. Figure 8: Medium resolution simulation results performed with [PITH_FULL_IMAGE:figures/full_fig_p009_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

26 extracted references · 24 canonical work pages

  1. [1]

    K., Linker, J

    Antiochos, S. K., Linker, J. A., Lionello, R., et al. 2013, in Multi-scale Physics in Coronal Heating and Solar Wind Acceleration, ed. D. Burgess, J. Drake, E. Marsch, R. von Steiger, M. Velli, & T. Zurbuchen, V ol. 38, 169–185

    work page 2013

  2. [2]

    N., Odstrcil, D., Pizzo, V

    Arge, C. N., Odstrcil, D., Pizzo, V . J., & Mayer, L. R. 2003, in American In- stitute of Physics Conference Series, V ol. 679, Solar Wind Ten, ed. M. Velli, R. Bruno, F. Malara, & B. Bucci, 190–193

    work page 2003

  3. [3]

    2024, Monthly Notices of the Royal As- tronomical Society, 529, 2399

    Bacchini, F., Ruan, W., & Keppens, R. 2024, Monthly Notices of the Royal As- tronomical Society, 529, 2399

    work page 2024

  4. [4]

    N., Odstrcil, D., Anderson, B

    Baker, D. N., Odstrcil, D., Anderson, B. J., et al. 2009, Journal of Geophysical Research (Space Physics), 114, A10101

    work page 2009

  5. [5]

    2022, A&A, 667, A133

    Baratashvili, T., Verbeke, C., Wijsen, N., & Poedts, S. 2022, A&A, 667, A133

    work page 2022

  6. [6]

    2012, Journal of Geophysical Research (Space Physics), 117, A08105

    Hayashi, K. 2012, Journal of Geophysical Research (Space Physics), 117, A08105

    work page 2012

  7. [7]

    2018, Space weather, 16, 1488 –1497

    Hill, F. 2018, Space weather, 16, 1488 –1497

    work page 2018

  8. [8]

    V ., Odstrcil, D., Yu, H

    Jackson, B. V ., Odstrcil, D., Yu, H. S., et al. 2015, Space Weather, 13, 104

    work page 2015

Show all 26 references
  1. [9]

    K., MacNeice, P

    Jian, L. K., MacNeice, P. J., Mays, M. L., et al. 2016, Space Weather, 14, 592

    2016

  2. [10]

    2023, A&A, 673, A66

    Keppens, R., Popescu Braileanu, B., Zhou, Y ., et al. 2023, A&A, 673, A66

    2023

  3. [11]

    K., Pogorelov, N

    Kim, T. K., Pogorelov, N. V ., Borovikov, S. N., et al. 2014, Journal of Geophys- ical Research (Space Physics), 119, 7981

    2014

  4. [12]

    O., Arge, C

    Lee, C. O., Arge, C. N., Odstrˇcil, D., et al. 2013, Sol. Phys., 285, 349

    2013

  5. [13]

    2021, Journal of Geophysical Research (Space Physics), 126, e28870

    Li, H., Feng, X., & Wei, F. 2021, Journal of Geophysical Research (Space Physics), 126, e28870

    2021

  6. [14]

    A., Caplan, R

    Linker, J. A., Caplan, R. M., Downs, C., et al. 2016, in Journal of Physics Con- ference Series, V ol. 719, Journal of Physics Conference Series (IOP), 012012

    2016

  7. [15]

    A., et al

    Lionello, R., Downs, C., Linker, J. A., et al. 2013, ApJ, 777, 76

    2013

  8. [16]

    2008, ApJ, 684, 1448

    Manchester, Ward B., I., V ourlidas, A., Tóth, G., et al. 2008, ApJ, 684, 1448

    2008

  9. [17]

    G., Lyon, J

    Merkin, V . G., Lyon, J. G., Lario, D., Arge, C. N., & Henney, C. J. 2016, Journal of Geophysical Research (Space Physics), 121, 2866

    2016

  10. [18]

    G., Lyon, J

    Merkin, V . G., Lyon, J. G., McGregor, S. L., & Pahud, D. M. 2011, Geo- phys. Res. Lett., 38, L14107

    2011

  11. [19]

    & Pizzo, V

    Odstrcil, D. & Pizzo, V . J. 2009, Sol. Phys., 259, 297 Odstrˇcil, D., Riley, P., & Zhao, X. P. 2004, J. Geophys. Res., 109, 2116

    2009

  12. [20]

    M., Merkin, V

    Pahud, D. M., Merkin, V . G., Arge, C. N., Hughes, W. J., & McGregor, S. M. 2012, Journal of Atmospheric and Solar-Terrestrial Physics, 83, 32

    2012

  13. [21]

    & Poedts, S

    Pomoell, J. & Poedts, S. 2018, Journal of Space Weather and Space Climate, 8, A35

    2018

  14. [22]

    & Vainio, R

    Pomoell, J. & Vainio, R. 2012, ApJ, 745, 151

    2012

  15. [23]

    P., & Keppens, R

    Porth, O., Xia, C., Hendrix, T., Moschou, S. P., & Keppens, R. 2014, ApJS, 214, 4

    2014

  16. [24]

    A., Lionello, R., & Mikic, Z

    Riley, P., Linker, J. A., Lionello, R., & Mikic, Z. 2012, Journal of Atmospheric and Solar-Terrestrial Physics, 83, 1

    2012

  17. [25]

    2022, A&A, 662, A50

    Verbeke, C., Baratashvili, T., & Poedts, S. 2022, A&A, 662, A50

    2022

  18. [26]

    2021, ApJ, 908, L26 Article number, page 10 of 10

    Wijsen, N., Samara, E., Aran, À., et al. 2021, ApJ, 908, L26 Article number, page 10 of 10

    2021

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.