Sensitivity of source sediment fingerprinting modelling to tracer selection methods
Bibliographic record
Abstract
Abstract. In a context of accelerated soil erosion and sediment supply to water bodies, sediment fingerprinting techniques have received an increasing interest in the last two decades. The selection of tracers is a particularly critical step for the subsequent accurate prediction of sediment source contributions. To select tracers, the most conventional approach is the so-called three-step method, although, more recently, the consensus method has also been proposed as an alternative. The outputs of these two approaches were compared in terms of identification of conservative properties, tracer selection, contribution modelling tendency and performance on a single dataset. As for the tree-step method, several range test criteria were compared, along with the impact of the discriminant function analysis (DFA). The dataset was composed of tracing properties analysed in soil (through the consideration of three potential sources; n = 56) and sediment core samples (n = 32). Soil and sediment samples were sieved to 63 µm and analysed for organic matter, elemental geochemistry and diffuse visible spectrometry. Virtual mixtures (n = 138) with known source proportions were generated in order to assess model accuracy of each tracer selection method. The Bayesian un-mixing model MixSIAR was used to predict source contributions on virtual mixtures and actual sediments. The different methods tested in the current research can be distributed into three groups according to their more or less restrictive identification of conservative properties, which were found to be associated with different sediment source contribution tendencies. The less restrictive selections of tracers were associated with a dominant and constant contribution of forests to sediment, whereas the most restrictive selections were associated with dominant and constant contributions of cropland to sediment. In contrast, intermediately restrictive selection of tracers led to more balanced contributions of both cropland and forest to sediment production. Virtual mixtures allowed to compute several evaluation metrics, which supported a better understanding of each tracer selection modelling accuracy. However, strong divergences were observed between the predicted contributions of virtual mixtures and the predicted sediment source contributions. These divergences may likely be attributed to the occurrence of a non-(fully) conservative behaviour of potential tracing properties during erosion, transport and deposition processes, which could not be reproduced when generated the virtual mixtures. Among the compared tracer selection methods, the three-step method using the mean ± SD and hinge range test criteria provided the most reliable tracer selection methods. In the future, it would be fundamental to generate more reliable metrics to assess conservativeness, to support more reliable modelling and more realistic virtual mixture generation to correctly evaluate modelling accuracy. These improvements may contribute to trustworthy sediment fingerprinting techniques for supporting efficient soil conservation and watershed management.
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 | 0.024 | 0.057 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.002 | 0.001 |
| Open science | 0.001 | 0.002 |
| Research integrity | 0.002 | 0.001 |
| 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".