QTQt: Beyond the User Manual

Alyssa Abbey, California State University-Long Beach; Mark Wildman, University of Glasgow; Kerry Gallagher, Université de Rennes; Kendra Murray, Idaho State University; Andrea Stevens-Goddard, Indiana University-Bloomington
Author Profile
Initial Publication Date: June 6, 2024

Summary

This module works through a series of prepared rock thermal history modeling activities using the freely available software package QTQt. In the first few activities, students will (i) learn how to build input data files, (ii) create forward models and synthetic data, and (iii) set-up and run inverse models and learn to evaluate the output (thermal history and data predictions) against a 'true solution'. Activities are specifically designed to hone students' intuition about how the decisions made during a model set-up directly influence the model outputs. Students will devise strategies for critically examining input parameters, model output results, and exploring alternative solutions (i.e., sensitivity testing). The exercises are deceptively simple and will appeal to beginning and experienced users alike, because they simultaneously offer an accessible entry-point for QTQt and guide the user through experiences that form the foundation of any robust modeling practice.

Share your modifications and improvements to this activity through the Community Contribution Tool »

Context

Audience

This module is designed for use in a graduate course or advanced undergraduate course for Geo/Earth Science students with some background in Geochronology and/or Thermochronology. This module includes short lecture sessions, small and large group discussions, and applied activities that may be completed in a lab, as a series of lab type sessions, or as homework sets. This module can also be used on its own as training for individual students learning about thermochronological modeling methods and techniques using QTQt.

Skills and concepts that students must have mastered

Students should have a basic understanding of various geochronological methods, and general knowledge of thermochronology techniques and limitations. Students should also have a good understanding of geological processes related to the interpretation of rock thermal histories (e.g., uplift, faulting, erosion, sedimentary burial).

How the activity is situated in the course

This module has been tested in a stand-alone workshop setting for graduate students and thermochronology professionals both familiar and unfamiliar with modeling practices. Parts of the module have also been used as a sequence of training exercises for undergraduate students doing independent research projects. The module could easily be inserted as part of a geo- or thermo-chronology course taught on a quarter or semester schedule. This module could be used to introduce and develop an intuitive understanding of a specific thermochronometric system (e.g. apatite (U-Th-Sm)/He or apatite fission track) following a theoretical introduction to thermochronology, and/or to learn how thermal history modeling techniques translate geochronological data into geological interpretations. These exercises will also give students a more fundamental insight into the practical challenges related to geocomputational modeling and results communication.

Goals

Content/concepts goals for this activity

- Thermal history modeling

- Becoming familiar with the QTQt program user interface and model outputs

- Creating data input files and sensible parameterisation of models

- Designing and executing forward and inverse models

Higher order thinking skills goals for this activity

- Recognize, record, and map key decision-making points during model design.

- Develop rigorous sensitivity tests for models by exploring alternative decision-making.

- Critically evaluate a series of competing models and/or models with different statistical inferences.

- Develop an intuitive understanding of the types of thermal histories that might be expected for a thermochronology data set.

- Develop an awareness of transparent reporting

- Practice presentation and communication of results and modeling protocols (akin to protocols and data collection in a lab environment).

Other skills goals for this activity

- Collaborate with peers who have varying degrees of expertise

- Practice model interpretation and results communication

Description and Teaching Materials

This module requires students to review materials before doing the group activities. There are four different activities with lectures, applications and discussions, and a follow-up homework or project activity. Version 5.7.1 is provided here as a downloadable zip file for both Mac and Windows machines. If you would like a newer version, please email Kerry Gallagher (kerry.gallagher@univ-rennes1.fr). Provided is a PowerPoint file that can be used to facilitate movement through the activities and discussions. The file includes notes on where parts can be shortened or extended to fit different instructional modes or formats. Also provided is a written tutorial packet with step-by-step instructions for students to follow when performing each of the different exercises.

  1. Before students begin creating input files and running models, there are a series of videos they should watch and papers to read to obtain useful background information about the methods, use, and mathematical approaches that underlie the QTQt program. The videos were presented and recorded (in 2021) by Kerry Gallagher (creator of QTQt) and are linked below in the resources section. Thesuggested papers include Wolf et al., 1998 - to become familiar with the non-uniqueness of thermochronology ages; Gallagher, 2012 - which introduces the QTQt program; and Abbey et al., 2023 - which exemplifies the types of tests that will be completed in the activities of this module. We also suggest keeping the QTQt user guide on hand (provided in the supporting materials section). 
  2. Exercise 1: Building input data files (~30-45 min depending on instructor familiarity with QTQt and class size). Students will use a data table to build an input data file, and practice reviewing an existing data file.
  3. Exercise 2: Forward modeling (~30 min to 1 hr depending on instructor familiarity with QTQt and class size). Students will practice drawing thermal histories, creating thermal history input files, loading thermal history input files, running forward models, plotting data predictions, and generating synthetic data.
  4. Exercise 3: Inverse modeling (~1 hour). Students will learn how to set thermal history constraints (e.g., inverse model priors, and MCMC parameters), run inversions, plot data predictions, and compare results from model runs with few iterations vs. many.
  5. Exercise 4: Testing User Decisions on Model Outputs (~1-4 hours depending on how many tests or groups are desired and how much time is spent on discussion of results). Students will test what happens to model outputs when modifying initial constraints, general priors, types of data, and data errors. We suggest these tests are done in small groups, but they can also be done individually. Small groups are assigned a series of tests to run for one of the 4 scenarios listed above. They assign a reporter and note taker who will report results and observations to the larger group after testing is completed. The testing is followed by group discussions about how and why model results changed under the assigned modifications.   
  6. Group discussion and demonstration about obtaining an inverse model result that produces the exact/true thermal history (~30 min). We created synthetic data from a known thermal history, and yet, because of the nature of the mathematical approaches followed in QTQt, we get model results that are similar but not exact. Having a discussion about why that is, will help build users' intuition when making interpretations about what model outputs mean. We have also provided an example of what kind and how much data is needed and what modeling decisions can be used to create a near exact replica of the true/known thermal history (in the PowerPoint). Adding some discussion about how, in real geologic scenarios, we do not know the true thermal history and thus justifying the modeling decisions that are made, will be a key component of this activity.
  7. Exercise 5 - A case study (2-4 hours). Students are given data and some geologic information. Either in-class (as a lab) or outside of class (maybe as a homework project), students should work through the process of building the input data files, deciding on general priors, constraint boxes, and MCMC parameters, and run some inverse models to test different questions about the decisions they make given the geologic information they have. We suggest students use the reporting template in Abbey et al., 2023 (supplemental tables) to keep track of each test's decisions, reasoning, and results. Following this up with a group discussion about the different choices the students made and the various tests they performed and why concludes the module (~30 min).

Teaching Notes and Tips

It is important to note that QTQt is a program created and modified by a sole person (Kerry Gallagher). The program code is not open sourced; however, Kerry is constantly revising and adding to QTQt to increase capabilities and functionality. Therefore, there are many versions of QTQt. The future versions include even more options for decision-making and customizing modeling choices. In addition to this, the program is a bit buggy. Sometimes the program will quit unexpectedly. This can happen from opening the wrong file type, from clicking buttons in the wrong order, or any number of things. We suggest informing students that this will happen and that they should practice not becoming frustrated. Many of these bugs are not noticeable until they happen and most are easy to avoid.

Before starting this module or any parts of it, we suggest that QTQt should be downloaded to the computer(s) of choice and opened to make sure it is working before beginning. This saves a lot of troubleshooting time in class. Please refer to the provided document with tips for installation problems.

This module highlights the exploratory nature of modeling and testing decisions made when modeling. There are no "correct" answers. As an instructor, we encourage you to facilitate open discussion about the nature of modeling and let students explore and justify choices made throughout the process. It will help; however, if you have tried running through the step-by-step instructions yourself, since there is a high chance the program will quit unexpectedly. If you have run into those difficulties yourself, you will be better acquainted with how to help students through such a scenario.
The provided PowerPoint is meant as a guide for staying on track with each activity, there are some added notes for instructors about what to discuss or prompt, and there are notes about cutting or adding information at different parts of each activity.

Please do not hesitate to message co-creator Alyssa Abbey (alyssa.abbey@csulb.edu) with questions, clarifications, stories successes, or ideas related to this teaching module.


Assessment

Assessment of this module is mostly formative. During each exercise there are built-in discussion questions that can be answered by individuals, small groups or as a whole class.

Assessment of the final activity (given as a lab or homework) should include an emphasis on the justification or reasoning behind each modeling decision that was made. Having students use the provided templates (supplemental tables in Abbey et al., 2023) to record those choices and results would be a good method. There could also be a written or oral component where students present their modeling results and back up their models with the reasoning for the decision they made.

References and Resources

There are five recorded videos that are important for this module. All can be accessed through YouTube.
QTQt_1: Inversion Concepts: https://www.youtube.com/watch?v=Nb5U9XIwN4o&t=27s
QTQt_2: Capacities and user interface: https://www.youtube.com/watch?v=cPPu00o_L24&t=36s
QTQt_3: Building a data file: https://www.youtube.com/watch?v=BkYltIQv_XE&t=1642s
QTQt_4: Forward Modelling: https://www.youtube.com/watch?v=h_GDW57Q0Rs&t=15s
QTQt_5: Inverse Modelling: https://www.youtube.com/watch?v=Aven7jAk9Ko