Bias and error in modelling thermochronometric data: resolving a potential increase in Plio-Pleistocene erosion rate
Bibliographic record
Abstract
Abstract. Thermochronometry provides one of few methods to quantify rock exhumation rate and history, including potential changes in exhumation rate. Thermochronometric ages can resolve rates, accelerations, and complex histories by exploiting different closure temperatures and path lengths using data distributed in elevation. We investigate how the resolution of an exhumation history is determined by the distribution of ages and their closure temperatures through an error analysis of the exhumation history problem. We define the sources of error, defined in terms of resolution, model error and methodological bias in the inverse method used by Herman et al. (2013) which combines data with different closure temperatures and elevations. The error analysis provides a series of tests addressing the various types of bias, including addressing criticism that there is a tendency of thermochronometric data to produce a false inference of faster erosion rates towards the present day because of a spatial correlation bias. Tests based on synthetic data demonstrate that the inverse method used by Herman et al. (2013) has no methodological or model bias towards increasing erosion rates. We do find significant resolution errors with sparse data, but these errors are not systematic, tending rather to leave inferred erosion rates at or near a Bayesian prior. To explain the difference in conclusions between our analysis and that of other work, we examine other approaches and find that previously published model tests contained an error in the geotherm calculation, resulting in an incorrect age prediction. Our reanalysis and interpretation show that the original results of Herman et al. (2013) are correctly calculated and presented, with no evidence for a systematic bias.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.023 | 0.127 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.003 |
| Science and technology studies | 0.001 | 0.002 |
| Scholarly communication | 0.003 | 0.003 |
| Open science | 0.001 | 0.002 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".