{"id":"W4292265076","doi":"10.3389/fenvs.2022.914493","title":"New techniques for old fires: Using deep learning to augment fire maps from the early satellite era","year":2022,"lang":"en","type":"article","venue":"Frontiers in Environmental Science","topic":"Fire effects on ecosystems","field":"Environmental Science","cited_by":5,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University","funders":"","keywords":"Satellite imagery; Remote sensing; Land cover; Deep learning; Multispectral image; Multispectral Scanner; Vegetation (pathology); Environmental science; Satellite; Environmental resource management; Disturbance (geology); Forest ecology; Convolutional neural network; Land use; Ecosystem; Geography; Cartography; Computer science; Ecology; Artificial intelligence; Geology","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001312296,0.0002604496,0.0002435747,0.00008154369,0.001186785,0.0001187452,0.001410916,0.00004977399,0.0005377032],"category_scores_gemma":[0.0000578341,0.000233647,0.00008851512,0.0007006795,0.0004116026,0.0004816099,0.001295129,0.0003729633,0.00005732178],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.003095886,"about_ca_system_score_gemma":0.00002731303,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00543631,"about_ca_topic_score_gemma":0.0001041206,"domain_scores_codex":[0.9967151,0.0001735213,0.0003572818,0.0009117123,0.001089593,0.0007528102],"domain_scores_gemma":[0.9989861,0.0001154527,0.000152393,0.0004960911,0.000001300534,0.0002486604],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.00007025139,0.00008454335,0.5413632,0.000003396636,0.000008742312,0.00001016623,0.004788673,0.01193023,0.04182313,0.000004286762,0.003119806,0.3967935],"study_design_scores_gemma":[0.0008949628,0.0008935031,0.6364486,0.00006954394,0.00003620383,0.00001735466,0.005437856,0.07161576,0.01399304,0.0007644201,0.2686941,0.001134666],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9867601,0.0004883463,0.009394629,0.0003557845,0.0009202267,0.001682681,0.00004478053,0.0000657989,0.0002876672],"genre_scores_gemma":[0.9291492,0.00004319655,0.06920777,0.0005097248,0.00009916056,0.0002686689,0.0000115771,0.00003793772,0.0006727428],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.3956589,"threshold_uncertainty_score":0.9527849,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.005362107164648982,"score_gpt":0.1996216053893604,"score_spread":0.1942594982247114,"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."}}