{"id":"W4408195974","doi":"10.1038/s41597-025-04715-4","title":"A Comprehensive Dataset of Surface Water Quality Spanning 1940-2023 for Empirical and ML Adopted Research","year":2025,"lang":"en","type":"article","venue":"Scientific Data","topic":"Water Quality and Pollution Assessment","field":"Environmental Science","cited_by":18,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Water quality; Quality (philosophy); Environmental science; Surface water; Data science; Computer science; Biology; Ecology; Environmental engineering; Physics","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001096957,0.0006254075,0.0005800823,0.002776922,0.0003657577,0.0006958436,0.001057325,0.0009791221,0.004887437],"category_scores_gemma":[0.004390985,0.0002141514,0.000825037,0.00856048,0.0003201453,0.0007853538,0.00120493,0.0008742703,0.00303905],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001181614,"about_ca_system_score_gemma":0.001683534,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.04204448,"about_ca_topic_score_gemma":0.0509225,"domain_scores_codex":[0.9991034,0.000157699,0.0001506,0.0002461337,0.000238764,0.0001033768],"domain_scores_gemma":[0.9970841,0.0007339667,0.0005630358,0.0005643394,0.0008783988,0.0001761371],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0003401678,0.0003244499,0.238373,0.001640608,0.0005705231,0.0004177258,0.0003085438,0.02099087,0.002163147,0.004132219,0.6702871,0.06045167],"study_design_scores_gemma":[0.0001455889,0.0001355362,0.4365044,0.0004967242,0.0001634054,0.0003276406,0.0008775764,0.01459864,0.002494909,0.004188466,0.5399461,0.0001211923],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"dataset","genre_gemma":"dataset","genre_scores_codex":[0.01838901,0.0003002123,0.00166376,0.0002932191,0.0000572412,0.00004770947,0.9772875,0.0002422212,0.001719144],"genre_scores_gemma":[0.03382837,0.0001608941,0.002478362,0.00008991953,0.00002197801,0.0001677077,0.9627218,0.0000296587,0.0005013269],"genre_candidate":"dataset","genre_consensus":"dataset","teacher_disagreement_score":0.04204448,"threshold_uncertainty_score":0.08359951,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.4412245770869852,"score_gpt":0.520185085393014,"score_spread":0.0789605083060288,"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."}}