{"id":"W4296482870","doi":"10.5194/hess-2022-334","title":"Hybrid forecasting: using statistics and machine learning to integrate predictions from dynamical models","year":2022,"lang":"en","type":"preprint","venue":"","topic":"Meteorological Phenomena and Simulations","field":"Earth and Planetary Sciences","cited_by":34,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université de Sherbrooke; University of Saskatchewan","funders":"U.S. Army Corps of Engineers; Natural Environment Research Council; Sight Research UK; Swiss Federal Institute for Forest, Snow and Landscape Research; Science Foundation Ireland; Canada First Research Excellence Fund; UK Research and Innovation","keywords":"Predictability; Downscaling; Computer science; Numerical weather prediction; Data assimilation; Machine learning; Climate model; Forcing (mathematics); Merge (version control); Ensemble forecasting; Forecast skill; Meteorology; Climatology; Artificial intelligence; Environmental science; Climate change; Precipitation; Mathematics; 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":"codex-gemma-dda1882f352a","candidate_categories":["insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.0002317587,0.0002122554,0.0002915753,0.0001167614,0.0004341217,0.0001320675,0.0001899854,0.00007395133,0.009964917],"category_scores_gemma":[0.0001601199,0.0001729934,0.00004520315,0.00008489197,0.0000502312,0.00007335196,0.0002450581,0.0008541605,0.00001088986],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00001694564,"about_ca_system_score_gemma":0.00005536555,"about_ca_topic_candidate":true,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01591476,"about_ca_topic_score_gemma":0.001695521,"domain_scores_codex":[0.9984813,0.0001888322,0.0003273542,0.0005044417,0.000245425,0.0002526596],"domain_scores_gemma":[0.999014,0.000429094,0.00009649066,0.0001740411,0.00003858166,0.0002477739],"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":[0.00002441034,0.000008596576,0.02063981,0.000007189111,0.00002767472,0.0000101471,0.00014759,0.9726578,0.000002022881,0.0002407937,0.00003211943,0.006201838],"study_design_scores_gemma":[0.00008113267,0.00009816962,0.008658952,0.000009011224,0.00004530979,0.000003335638,0.00007912706,0.9177642,2.964926e-7,0.07278608,0.0002895835,0.0001847815],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.5425897,0.0001426543,0.4479572,0.00006113326,0.0002821595,0.0002198275,0.005637064,0.00007336789,0.003036941],"genre_scores_gemma":[0.9083325,0.00002882344,0.08651274,0.0001364389,0.00007316186,0.000003317241,0.004544775,0.000006580418,0.0003617026],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.3657428,"threshold_uncertainty_score":0.9909401,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.09412237713857317,"score_gpt":0.2595672813664733,"score_spread":0.1654449042279001,"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."}}