MétaCan
Menu
Back to cohort
Record W4415989079 · doi:10.1144/esss2024-005

Quantifying the limited explanatory power of traditional classified geological maps

2025· article· en· W4415989079 on OpenAlex
Charlie Kirkwood

Why this work is in the frame

A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.

affAt least one author lists a Canadian institution in the pinned OpenAlex snapshot.

Bibliographic record

VenueEarth Science Systems and Society · 2025
Typearticle
Languageen
FieldComputer Science
TopicGeochemistry and Geologic Mapping
Canadian institutionsArtificial Intelligence in Medicine (Canada)
Fundersnot available
KeywordsGeologic mapExplanatory powerGeological surveyRelation (database)Bayesian probabilityBedrockVariance (accounting)Regional geologyClass (philosophy)

Abstract

fetched live from OpenAlex

Since William Smith published his ‘Delineation of the Strata of England and Wales with part of Scotland’ in 1815, it has been the accepted practice to map geology as a collection of discrete polygons (or volumes in 3D), each representing a parcel of rock that is sufficiently self-similar in terms of its geological properties to class as a ‘unit’. However, the properties that are used to define these units, such as age, composition and texture, are themselves continuous variables more suited to modelling by regression than by classification. As such, the discrete nature of the traditional classified geological map has limited power to explain real observed values of geological properties, which vary on continuous scales, resulting in a disconnect between our geological maps and reality. In this study, we quantify the explanatory power of a traditional geological map – the British Geological Survey's 1 : 625K bedrock geology of the UK – in relation to observations of major element chemical composition, and compare this with the explanatory power provided by a Bayesian deep learning regression-based approach to ‘properties-first’ geological mapping. We find that, for our selection of elements, the traditional geological map explains between 57 and 66% of their variance, and the Bayesian deep learning approach explains between 66 and 75% of their variance, almost 10% higher explained variance than the traditional classified geological map, which equates to progress of between a fifth and quarter of the way towards achieving the ‘ultimate’ perfectly informative geological map. We discuss the implications of the traditional classified geological map's limited explanatory power and how the inherent constraint of its discrete classified representation should reasonably lead us to pursue geological mapping techniques that are free of this limitation, i.e. which are based on regression rather than classification.

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.

Full frame distilled prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.

metaresearch head score (Codex)0.001
metaresearch head score (Gemma)0.000
Version: codex-gemma-dda1882f352aValidation status: machine_predicted_unvalidated
Candidate categoriesnone
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Theoretical or conceptual · Consensus signal: none
GenreCandidate signal: Empirical · Consensus signal: Empirical
Teacher disagreement score0.957
Threshold uncertainty score0.434

Codex and Gemma teacher scores by category

CategoryCodexGemma
Metaresearch0.0010.000
Meta-epidemiology (narrow)0.0000.000
Meta-epidemiology (broad)0.0000.000
Bibliometrics0.0000.001
Science and technology studies0.0010.001
Scholarly communication0.0000.000
Open science0.0010.000
Research integrity0.0000.000
Insufficient payload (model declined to judge)0.0000.000

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.

Opus teacher head0.054
GPT teacher head0.249
Teacher spread0.195 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it