| 08 Aug 2025
Technical note: Geodynamic Thermochronology (GDTchron) – A Python package to calculate low-temperature thermochronometric ages from geodynamic numerical models
Abstract. Low-temperature thermochronology provides a powerful means of extracting quantitative information on the thermal evolution of different tectonic settings from rocks exposed at the surface of the Earth. Geodynamic numerical models enable tracking the entire thermal structure of simulated tectonic settings throughout their evolution. Despite the highly complementary nature of these two approaches, few geodynamic modeling studies have used the thermal information in models to predict thermochronometric ages as a means of comparing model results with observational data. Here, we present Geodynamic Thermochronology (GDTchron): an open-source Python package designed to forward model large numbers of low-temperature thermochronometric ages from time-temperature paths output by geodynamic numerical models. This package uses existing techniques to estimate apatite (U-Th)/He, apatite fission track, and zircon (U-Th)/He ages from time-temperature paths in a parallelized workflow that enables faster computation on multicore processors and high-performance computing systems. It is designed to extract the temperature of many selected particles over multiple timesteps. Our workflow is built on typical output files from geodynamic models containing particle location, time, and temperature, and we use an interpolation scheme to allow new particles to inherit the thermal histories of their nearest neighbors. GDTchron can be applied to any tectonic setting, though for results to be comparable to nature, geodynamic models should account for erosion and sedimentation. We demonstrate the functionality of this software with a highly simplified geodynamic model of uplift and a more complicated model of rift-inversion orogenesis with the aim of encouraging community participation in broadening future development.
General comments
The authors present a new, open-source Python package designed to forward model large numbers of low-temperature thermochronometric ages from time-temperature paths generated by geodynamic numerical models. The manuscript, and Python package, comprise timely and potentially substantial contributions to the fields of geochronology and geodynamics – as the authors mention in the paper, advances in geodynamic numerical modeling have enabled increasingly high-resolution simulations of geological and thermal events across various tectonic settings and thermochronology data provide quantitative constraints against which such models can be validated. The GDTchron Python package presented here expands on thermokinematic modeling approaches (which predict time-temperature histories and low-temperature thermochronology ages based on deformation kinematics and topographic evolution but do not directly incorporate dynamics; Braun, 2003; Braun et al., 2012; Almendral et al., 2015) and previous strategies for integrating time-temperature information from geodynamic models with thermochronology data (which has involved exporting t-T paths from geodynamic models and importing them into thermal history modeling software like HeFTy or QTQt; Ketcham, 2005; Gallagher, 2012) in that it allows for extraction of time-temperature histories, and thermochronology age predictions, for a large number of particles from multiple steps in a geodynamic model.
The paper is well written, the accompanying Python notebooks are well documented/commented and readily accessible via Github, and the figures are generally illustrative. However, this research has several areas that I consider in need of major improvements before manuscript publication and widespread implementation of the Python package by users in the geodynamics and thermochronology communities:
Specific comments
Technical corrections
Line 54: Specify that the simplified model involves block uplift here?
Line 55: Because routines incorporating erosion and sedimentation are so important in predicting cooling ages for low-T thermochronometers, I strongly recommend providing more information about what these routines were, & what geological or other evidence supports their use, here in the intro in addition to later in the paper.
Line 109 / Section 3.1: What kind of computational resources are required to run GDTchron? Can you provide more information here regarding the types of scenarios that might run on a personal computer/workstation, versus HPC? (I recognize that related information is documented within the Github & other package resources, but would be valuable to include for readers in this section of the paper too)
Line 141: As a reviewer with expertise in thermal history and structural/thermokinematics modeling but no hands-on experience with geodynamic software like that listed above, I am curious what common inputs and outputs to these software are, particularly as they relate to the temperature field, partical motion, and the development/decay of topography. Even though your target audience may already be familiar with these programs, I recommend including a more comprehensive background/context section that helps frame geodynamic model capabilities, and limitations, for a broader audience. E.g., how is topography typically handled in these systems, and what are recommendations/considerations for evaluating incorporation of topographic change and sedimentation in geodynamic models? In addition, there may be readers with expertise in thermochronology, but not ASPECT etc. modeling, who would like to use GDTchron to extract t-T paths and predict cooling ages based on published geodynamic models to help inform their sampling strategies. Providing a little more context for the geodynamic modeling component would help clarify the intended workflow and capabilities of the GDTchron package.
Line 169: Choose a different word than “flow” when discussing how material exits the top of the box? “Flow” has implications for rheology/deformation style that do not necessarily apply here.
Line 170: By “temperature structure of the box remains constant,” do you mean that there is no advection or change in isotherms due to particle uplift & exhumation? It appears so, from the model diagrams? If not, please consider revising this simple example to allow for isotherm advection, which will produce a more geologically accurate case study?
Line 170: Specify "particle / rock uplift" or "exhumation" here, as there is no surface/topographic uplift in this scenario that maintains flat topography (e.g., has complete erosional efficiency).
Lines 176-177: Language related to Fig. 5 here is a little confusing – could be interpreted to mean that the timesteps were generated with GDTchron, but they were defined in ASPECT, right? Then the outputs of that model were processed in GDTchron?
Lines 179-180, sentence starting with “Ages young with depth…”: Phrasing makes it seem like 0 Ma is reached spatially below the PRZ/PAZs, but the unit is time, not space. Change to "Ages young with depth through the partial retention/annealing zones of each thermochronometric system. At the base of these zones, cooling ages are 0 Ma because diffusion and annealing outpace He and fission track production" or similar?
Line 190 / Fig. 5: I strongly recommend devising a way to show the movement/change in particle positions in this figure - as vectors, a box with no fill showing before/after positions, positions of partial annealing/retention zones before and after uplift, etc. For Y-axis for Figs. 4 and 5, I strongly recommend flipping the axis such that 0 km is at the surface, and values increase with depth (or negative elevation). This recommendation uses the same perspective as structural cross sections and other geodynamic models, and I think would make both thermal structure and predicted cooling age plots more intuitive for readers and potential GDTchron package users.
Lines 194-196, sentence starting with “One could explore the effects…”: Predicted cooling ages for all of these scenarios also depend on the mechanism for incorporating or approximating topographic evolution – which may have just as much impact on the distribution of predicted cooling ages as the aforementioned factors, especially for low-T thermochron data in regions of relatively slow rock uplift/exhumation (Gilmore et al., 2018; Ketcham, 2025; & references therein).
Lines 200-204: Completely agreed – though I think it would be helpful to articulate how these factors are incorporated into geodynamic models to help educate readers and ensure potential GDTchron users make effective choices when coupling geodynamic models and interpretations of thermochronology data.
Lines 214-215, sentence starting with “Although thermokinematic models…”: Consider including a brief discussion of the similarities & differences of each approach, e.g., Pecube versus GDTchron? What is each good at (e.g., Pecube incorporates a sophisticated framework for estimating topographic evolution, Pecube-D tracks thermal histories of particles involving large amounts of horizontal shortening/extension and complex faulting, statement on how these relate to the conditions of most geodynamic models?)
Line 220 / Fig. 6: Please clarify that ages are relative to the start time/duration of the model. Consider flipping Y axes for both the model grid (so that 0 km is at Earth's surface and depth/negative elevation increase downwards) and the Surface Ages plots (so that 0 Ma is at the bottom, and older ages are on top - this is the framework commonly used for Pecube-D and other thermokinematic models). What do you mean by "Maximum Age (Ma)" as Y-axis label on bottom charts? Seems like these are just predicted cooling ages? Please consider adding some vertical exaggeration to the middle plots, and perhaps intermediate ticks/values to the Plastic Strain and AHe Age gradient keys – predicted cooling ages only use a third of the space, and it is challenging to read these gradient values without zooming in 200+% into the figure.
Line 229, sentence starting with “This 2D model…”: Which model? That in Vasey et al. (2024) using hillslope diffusion to approximate surface processes, or the revised model in this study that uses Fastscape? Here, it would be valuable to use this example, and the changes made to "provide more realistic surface processes", as a framework for discussing what changes were made to the mechanisms of approximating topographic evolution and their influence on predicted cooling ages.
Line 231, sentence starting with “At 16 Myr…”: Time reference is unclear here - do you mean after 16 Myr into the model run, or at 16 Ma? At the end of the sentence, it would be helpful to add “… with a total model run time of 36 Myr" or similar.
Lines 236-237, “resulting in slightly younger AHe (~10 Ma) and AFT (~14 Ma) ages…”: Here, can also emphasize that these predicted ages are partially reset with respect to inherited model age?
Line 240: Update “…lithospheric thickening and uplift” to “lithospheric thickening, uplift, and erosion”
Lines 259-261, “Implementing these alternative kinetic models…”: Why not incorporate the most widely used kinetic models in this first release, especially if kinetic models are styled after those used in Ketcham (2005)? Doing so would significantly expand the potential for comparisons between predictions based on geodynamic models and thermochronology data for specific study areas.
Lines 263-265, “Importantly, the ages produced when applying outputs from geodynamic models are dependent on parameters that would be highly variable among samples within real systems (e.g., U and Th concentrations, grain size, Dpar)…”: Here, viable solutions might include incorporating values based on best practices (e.g., for grain sizes for He analyses) or values based on standards, then allowing for users to adjust GDTchron sample parameters to best match thermochronology data of interest? This suggestion is inspired by similar options in thermokinematic modeling programs including PecubeGUI (Bernard et al., 2025).
References not included in the manuscript:
Almendral, A., Robles, W., Parra, M., Mora, A., Ketcham, R. A., & Raghib, M. (2015). FetKin: Coupling kinematic restorations and temperature to predict thrusting, exhumation histories, and thermochronometric ages. AAPG Bulletin, 99(8), 1557-1573.
Bernard, M., van der Beek, P., Colleps, C., & Amalberti, J. (2022, May). PecubeGUI: a new graphical user interface for Pecube, introduction and sample-specific predictions of apatite (U-Th)/He and 4He/3He data in the Rhone valley, Switzerland. In EGU General Assembly Conference Abstracts (pp. EGU22-2277)., 524–525, 1–28, https://doi.org/10.1016/j.tecto.2011.12.035, 2012.
Gilmore, M. E., McQuarrie, N., Eizenhöfer, P. R., & Ehlers, T. A. (2018). Testing the effects of topography, geometry, and kinematics on modeled thermochronometer cooling ages in the eastern Bhutan Himalaya. Solid Earth, 9(3), 599-627.
Ketcham, R. A. (2025). Incorporating topographic deflection effects into thermal history modelling. Geochronology, 7(3), 449-458.