Evaluation of Filtering Methods for Hydrograph Separation in Small Agricultural Watersheds in Québec, Canada
Bibliographic record
Abstract
Highlights Agricultural hydrology is complex due to the management of surface and subsurface flow to increase productivity. This study provides an interpretation of hydrological functioning, using a geochemical tracer (electrical conductivity) as a reference method, for hydrograph separation and evaluation of filtering methods. Filtering method efficiency must be interpreted according to season, year, watershed relief, and management practices. Routine application of basic filtering concepts is not sufficient to address the heterogeneity of hydrological processes in agricultural watersheds. Abstract. Streamflow hydrographs summarize the behavior of watersheds. Their separation into quick and slow components requires hydrological knowledge of the specific drainage area. To better understand the hydrological response of 14 small agricultural watersheds in Québec, Canada, covering different physiographic attributes ranging from lowlands to hilly and steep landscapes, streamflow electrical conductivity was used as a geochemical tracer. These agricultural watersheds have undergone significant management practices, including artificial drainage. The objective of this research was to evaluate the performance of existing automated filter methods for hydrograph separation (BFLOW, UKIH, PART, FIXED, SLIDE, LOCMIN, and Eckhardt). The geochemical method was used as a reference for comparison with the filter methods. Comparison of the slow flow estimates from non-calibrated filters, using a MANOVA model, showed that the filter performance increased under conditions with high contributions of quick runoff to the stream, such as during snowmelt (spring season), during heavy precipitation, and in subwatersheds with landscape conditions more prone to quick runoff. However, filter performance decreased as hydrological processes predisposed more flow to slower pathways, typically in summer and fall, as well as in lowland landscapes generally associated with high rates of tile drainage rather than in hilly and steep relief. Underlying the filter assumptions is the classic concept of a rainfall event with quick runoff as the main source of the drainage area response. Thus, slow flow is associated with a low threshold response. Eckhardt filter simulations were in good agreement with the geochemical method after calibration, based on model statistical measures (R, NSE, and PBIAS). However, larger errors were associated with higher flow values. The slow flow overestimations were more pronounced during periods of extreme events, i.e., spring runoff and heavy precipitation. The linear concept of the Eckhardt filter yields no information on slow flow response behavior that could be useful in capturing its temporal variability. Because the routing of water has been managed to improve agricultural productivity, these hydrological modifications resulted in a more complex slow flow response. The performance of filtering methods is thus affected. Therefore, simplifications of filter assumptions are less likely to provide more effective estimates of slow flow. Furthermore, given the heterogeneity of hydrological processes due to seasonal climatic characteristics, the routine application of basic filter concepts is not sufficient to address the variable nature of the hydrological response. The variability scale of geochemical separation, from regional (agro-climatic) to local (adjacent watersheds), proved that it is always relevant to have adequate separation. However, the validation of filters without a tracer is limited and almost unsuitable for these agricultural watersheds. Keywords: Agricultural watershed, Artificial drainage, Electrical conductivity, Filtering method, Geochemical method, Hydrograph separation, MANOVA, Quick flow, Slow flow, Tile drainage.
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 | 0.001 | 0.004 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.002 | 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".