{"id":"W4313814630","doi":"10.5194/hess-27-139-2023","title":"Continuous streamflow prediction in ungauged basins: long short-term memory neural networks clearly outperform traditional hydrological models","year":2023,"lang":"en","type":"article","venue":"Hydrology and earth system sciences","topic":"Hydrology and Watershed Management Studies","field":"Environmental Science","cited_by":217,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo; École de Technologie Supérieure","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Streamflow; Robustness (evolution); Computer science; Artificial neural network; Hydrological modelling; Data set; Hyperparameter; Drainage basin; Machine learning; Data mining; Artificial intelligence; Climatology; Geology; Cartography; Geography","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":false,"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":[0.001233236,0.0004614723,0.0002957816,0.0003368763,0.0001700825,0.0007723508,0.0003445571,0.0004705175,0.0006991439],"category_scores_gemma":[0.003136392,0.0001484818,0.0002420163,0.000379384,0.0003126355,0.001511697,0.0004738755,0.000454373,0.0001151433],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006275412,"about_ca_system_score_gemma":0.0004626499,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01046478,"about_ca_topic_score_gemma":0.01585694,"domain_scores_codex":[0.9997849,0.00007935936,0.00001625307,0.00005960208,0.00002990403,0.00002998336],"domain_scores_gemma":[0.9992526,0.0004768622,0.00006203138,0.00005057797,0.0001244278,0.00003338316],"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":[0.0006748409,0.0001852766,0.06396233,0.0001219056,0.0003116633,0.0002098055,0.0001936591,0.7165189,0.01027197,0.001469465,0.001454856,0.2046254],"study_design_scores_gemma":[0.000008626783,0.00006035716,0.009065186,0.0000131835,0.00002102526,0.000009242838,0.00004036059,0.9881055,0.001701228,0.0008135269,0.000154,0.000007727063],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9790988,0.0005470272,0.0183189,0.0002644504,0.00003735455,0.00001112409,0.00009377508,0.0001849584,0.001443731],"genre_scores_gemma":[0.9967859,0.0000602471,0.002856517,0.00003026815,0.000004290302,0.000003250219,0.00005130739,0.000005410967,0.0002028507],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.01046478,"threshold_uncertainty_score":0.02080774,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03252885500243047,"score_gpt":0.2155571140919531,"score_spread":0.1830282590895226,"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."}}