{"id":"W7117166872","doi":"10.1016/j.wroa.2025.100475","title":"Using “big data” and non-linear machine learning to infer groundwater contamination mechanisms across a spatially extensive, geologically heterogeneous region","year":2025,"lang":"en","type":"article","venue":"Water Research X","topic":"Fecal contamination and water quality","field":"Environmental Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Public Health Ontario; Queen's University","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Aquifer; Groundwater; Hydrology (agriculture); Contamination; Groundwater recharge; Water quality; Hydrogeology; Fecal coliform","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001696129,0.0009052628,0.0004594539,0.001002353,0.0003972843,0.00130186,0.00105754,0.000713693,0.001021351],"category_scores_gemma":[0.004782834,0.0004154296,0.001409813,0.001249345,0.000773365,0.0008795839,0.001000262,0.0006990613,0.0001867195],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00227791,"about_ca_system_score_gemma":0.002003963,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.1442961,"about_ca_topic_score_gemma":0.1735317,"domain_scores_codex":[0.9994965,0.0002239922,0.00002936561,0.0001358531,0.00005663609,0.00005782026],"domain_scores_gemma":[0.9975672,0.001592156,0.0003258225,0.0002161271,0.0002054568,0.00009326232],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0002120496,0.0001531091,0.2668666,0.000267823,0.0009424071,0.0002862554,0.0002720238,0.6870116,0.001301507,0.003856767,0.001922332,0.03690758],"study_design_scores_gemma":[0.00001978748,0.00006454506,0.05558311,0.0000371969,0.0001057957,0.00003863465,0.0002253299,0.9353622,0.0005671118,0.006253178,0.001714302,0.00002883288],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8276025,0.001046048,0.1578194,0.00212548,0.00008951476,0.0001736613,0.00692231,0.0009673593,0.003253704],"genre_scores_gemma":[0.9729148,0.0002904888,0.02278543,0.0001088286,0.00002148402,0.00007073899,0.002826115,0.00003519703,0.0009470439],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.1442961,"threshold_uncertainty_score":0.2869124,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1654134129102717,"score_gpt":0.3896354168105627,"score_spread":0.2242220039002911,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}