Investigation and integration of spatial analyses in benthic habitat mapping with application to nearshore Arctic environments
Bibliographic record
Abstract
The field of benthic habitat mapping has entered an era of automated statistical methods that have increased the capacity to produce maps as marine management tools. Spurred by a confluence of advances in acoustic remote sensing, open-source statistical tools, GIS, and computing power, these methods facilitate quick and objective mapping of habitats and physical seabed characteristics. Their performance and accessibility have led to widespread uptake, yet key spatial issues associated with these methods have not fully translated into the benthic habitat mapping workflow. Towards establishing “best practices”, this thesis explores the application of several spatial concepts to benthic habitat mapping using three Canadian Arctic case studies. Relationships between seabed morphology and benthic habitats are well-established. Though recognized as a critical element in the field of geomorphometry, the scale dependence of these relationships is commonly neglected in habitat mapping. Chapter 2 provides evidence of the scale dependence of benthic terrain variables and demonstrates methods for testing and selecting from among many variables and scales for modelling the distribution of sediment grain size near Qikiqtarjuaq, Nunavut. Given challenges associated with marine data collection that are pronounced in the Arctic, benthic habitat maps commonly utilize multi-year and multisource datasets. Despite apparent advantages, there can be substantial challenges associated with the compatibility and spatial properties of such data. Chapter 3 demonstrates that spatially autocorrelated samples are likely to inflate estimates of predictive performance and uses a spatial resampling strategy to estimate and correct for inflation in a multi-model Arctic clam habitat map near Qikiqtarjuaq, Nunavut. Classified seabed maps are a common requirement for marine management and one of two broad approaches are often selected to produce them. Chapter 4 examines differences between classification and continuous modelling approaches in a spatial context to produce classified seabed sediment maps for inner Frobisher Bay, Nunavut. Non-spatial methods failed to indicate whether models could extrapolate to unsampled areas, which was a requirement for this study. When evaluated in a spatial context, the qualities of the classification approach made it more suitable, which was a function of ground-truth dataset characteristics and the predictive goals of the model. Non-spatial techniques may be appropriate for interpolation, but the ability to extrapolate needs to be examined in a spatial context.
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 | 0.005 | 0.018 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.001 |
| Bibliometrics | 0.003 | 0.007 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.005 | 0.002 |
| Open science | 0.001 | 0.002 |
| Research integrity | 0.000 | 0.001 |
| 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".