2026-10-08T09:41:23Z https://api.figshare.com/v2/oai

oai:figshare.com:article/33716494 2026-09-13T20:09:52Z category_26965 category_26971 category_26968 category_26935 category_26947 item_type_3 month_year_09_2026

When Does Water-Quality Prediction Generalize? An Observation–Process–Deployment Framework for Large-Sample River Modelling: code, frozen manifests and reproducibility materials Addilat Abdul Mumin (24764773) Groundwater quality processes and contaminated land assessment Pollution and contamination not elsewhere classified Surface water quality processes and contaminated sediment assessment Environmental assessment and monitoring Natural resource management water quality environmental modelling machine learning generalization domain shift river monitoring uncertainty quantification conformal prediction local adaptation Caravan-Qual <p dir="ltr">Reproducibility materials for the study “When Does Water-Quality Prediction Generalize? An Observation–Process–Deployment Framework for Large-Sample River Modelling”. The release contains the staged audit and modelling code, frozen validation/adaptation manifests, robustness and uncertainty analyses, manuscript-facing result tables, publication figures, final experiment gates, and the H3 environmental deployment distance audit. Raw Caravan-Qual data are not redistributed; users should obtain Caravan-Qual v1.1 from DOI 10.5281/zenodo.21337532. The repository evaluates random, temporal, unseen-station, unseen-catchment and held-out-climate-domain validation across five river-water-quality constituents and quantifies progressive local adaptation.</p> 2026-09-13T20:09:52Z Dataset Dataset 10.6084/m9.figshare.33716494.v1 https://figshare.com/articles/dataset/When_Does_Water-Quality_Prediction_Generalize_An_Observation_Process_Deployment_Framework_for_Large-Sample_River_Modelling_code_frozen_manifests_and_reproducibility_materials/33716494 CC BY 4.0