{"id":"W4297372102","doi":"10.5194/amt-15-5497-2022","title":"A new machine-learning-based analysis for improving satellite-retrieved atmospheric composition data: OMI SO <sub>2</sub> as an example","year":2022,"lang":"en","type":"article","venue":"Atmospheric measurement techniques","topic":"Atmospheric Ozone and Climate","field":"Earth and Planetary Sciences","cited_by":12,"is_retracted":false,"has_abstract":true,"ca_institutions":"Environment and Climate Change Canada","funders":"Earth Sciences Division; Goddard Space Flight Center; National Aeronautics and Space Administration","keywords":"Zenith; Pixel; Satellite; Solar zenith angle; Remote sensing; Environmental science; Noise (video); Data set; Ground truth; Daytime; Computer science; Meteorology; Artificial intelligence; Image (mathematics); Atmospheric sciences; Physics; Geography","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":["metaepi_narrow","sts","insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.002753311,0.0005674477,0.0007423383,0.000008680384,0.001502267,0.0003564248,0.001231833,0.0001474407,0.001981422],"category_scores_gemma":[0.00008949007,0.0005646425,0.0003209707,0.002122321,0.00009102005,0.0006900691,0.0001570678,0.0004949865,0.00003305803],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001651002,"about_ca_system_score_gemma":0.0005977018,"about_ca_topic_candidate":true,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0286432,"about_ca_topic_score_gemma":0.004009486,"domain_scores_codex":[0.9948273,0.0005425725,0.0008173003,0.00138228,0.001551053,0.0008794802],"domain_scores_gemma":[0.9972629,0.0001696909,0.0006103075,0.001311567,0.0002315999,0.0004139625],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.001142899,0.0003690635,0.1818404,0.0001314499,0.0007947162,0.00003822274,0.000431387,0.04299318,0.01491581,0.00004656451,0.001066424,0.7562299],"study_design_scores_gemma":[0.001627449,0.004010503,0.02761695,0.00002839126,0.001833163,0.00002139836,0.0006890243,0.864691,0.01710226,0.0005054964,0.08017261,0.001701781],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3695056,0.003934482,0.6210099,0.000286872,0.0003441733,0.002341338,0.0002210737,0.001788415,0.0005680566],"genre_scores_gemma":[0.8478287,0.0001444867,0.1476046,0.0007706344,0.0001552397,0.00009774573,0.003274369,0.0000488983,0.00007525543],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.8216978,"threshold_uncertainty_score":0.9997976,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04171867334665631,"score_gpt":0.240235481346631,"score_spread":0.1985168079999747,"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."}}