2026-10-09T20:18:23Z https://api.figshare.com/v2/oai

oai:figshare.com:article/34037813 2026-10-01T02:19:26Z category_1 category_272 category_873 category_39 category_931 portal_87 item_type_6 month_year_10_2026

A rapid-updating method for anthropogenic NO_x emissions based on convolutional neural networks and TROPOMI NO₂ observations Yucong Zhang (7342424) Steffen Beirle (7810742) Liangyun Liu (3162909) Leon Kuhn (25148096) Shanshan Du (780209) Liping Lei (161053) Thomas Wagner (95264) Biophysics Environmental Sciences not elsewhere classified Chemical Sciences not elsewhere classified Ecology Information Systems not elsewhere classified NOx emissions TROPOMI NO2 emission inventory deep learning CNN <p>Nitrogen oxides (NO_<i>x</i> = NO + NO₂) are air pollutants primarily emitted from anthropogenic sources, however, bottom-up anthropogenic NO_<i>x</i> emission inventories often suffer from update delays. The TROPOspheric Monitoring Instrument (TROPOMI) provides near-real-time, high-resolution NO₂ column densities, offering new opportunities for timely NO_<i>x</i> emission prediction. Given the short lifetime of NO_<i>x</i> in the lower troposphere, local emissions can influence observations over distances of ~100 km. In this study, we develop a convolutional neural network to predict anthropogenic NO_<i>x</i> emissions from TROPOMI NO₂ observations. The model operates at a monthly temporal resolution and 0.1 ° × 0.1 ° spatial resolution over Europe and the continental United States, using TROPOMI NO₂ columns and ERA5 wind fields as inputs, with EDGARv8.1 emissions as training targets. Trained on 2019–2020 data, the model achieves an R² of 0.958 and an RMSE of 8.194 Mg/month/cell on the test set. It demonstrates strong temporal generalisation, with an average R² of 0.927 at the grid scale and 0.966 at the national scale during 2021–2022, although spatial generalisation remains limited. The model is applied to extend gridded NO_<i>x</i> emissions to December 2025. Overall, the proposed satellite-driven deep learning approach enables accurate, high-resolution, and near-real-time updates of anthropogenic NO_<i>x</i> emissions.</p> 2026-10-01T02:19:26Z Text Journal contribution 10.6084/m9.figshare.34037813.v1 https://figshare.com/articles/journal_contribution/A_rapid-updating_method_for_anthropogenic_NO_sub_x_sub_emissions_based_on_convolutional_neural_networks_and_TROPOMI_NO_observations/34037813 CC BY 4.0