{"id":"W4415700244","doi":"10.1080/02626667.2025.2581264","title":"Enhancing long-term water quality forecasting with a hybrid deep-learning approach integrating MODWT, CNN, and GRU","year":2025,"lang":"en","type":"article","venue":"Hydrological Sciences Journal","topic":"Hydrological Forecasting Using AI","field":"Environmental Science","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université du Québec en Abitibi-Témiscamingue; University of Waterloo","funders":"","keywords":"Quality (philosophy); Water quality; Automation; Key (lock); Work (physics)","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["sts"],"consensus_categories":[],"category_scores_codex":[0.004664246,0.0003136312,0.000405111,0.0001002512,0.002180696,0.0004977396,0.0005953124,0.0001230701,0.0005639772],"category_scores_gemma":[0.0006341566,0.0001598581,0.00009247115,0.0005011605,0.001618726,0.0004880836,0.0006277132,0.001134785,0.00002533965],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001599752,"about_ca_system_score_gemma":0.00002116446,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0000593349,"about_ca_topic_score_gemma":0.00005054718,"domain_scores_codex":[0.9963612,0.000527208,0.0006101532,0.000795167,0.0006826721,0.001023583],"domain_scores_gemma":[0.9989908,0.0003146383,0.000246789,0.0001411989,0.00002622729,0.0002803271],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0001534904,0.0002087054,0.721992,0.00003728152,0.00003794509,0.0002228752,0.0009517297,0.2091608,0.01988498,0.0002568807,0.00002109677,0.04707218],"study_design_scores_gemma":[0.001524175,0.00281141,0.150883,0.0003561928,0.0001038304,0.004481706,0.0002980818,0.8187518,0.005738251,0.01350833,0.0003015223,0.001241727],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9291245,0.00007419841,0.0534215,0.0004552169,0.00008432479,0.0001519099,2.962592e-7,0.00009041846,0.01659763],"genre_scores_gemma":[0.9814406,0.00001396942,0.01774785,0.0005189701,0.00007027641,0.00001008557,0.00000180165,0.000008762177,0.0001876517],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.6095909,"threshold_uncertainty_score":0.9991183,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03420564455226331,"score_gpt":0.2756350293391229,"score_spread":0.2414293847868595,"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."}}