{"id":1488,"date":"2024-10-10T14:10:58","date_gmt":"2024-10-10T12:10:58","guid":{"rendered":"https:\/\/wordpress-test.app.u-pariscite.fr\/diip\/?p=1488"},"modified":"2024-10-25T11:19:20","modified_gmt":"2024-10-25T09:19:20","slug":"estimation-of-pollutant-emissions-from-remote-sensing-data-and-deep-learning-esperel","status":"publish","type":"post","link":"https:\/\/wordpress-test.app.u-pariscite.fr\/diip\/estimation-of-pollutant-emissions-from-remote-sensing-data-and-deep-learning-esperel\/","title":{"rendered":"EStimation of Pollutant Emissions from REmote sensing data and deep Learning (ESPEREL)"},"content":{"rendered":"

[et_pb_section fb_built=”1″ _builder_version=”3.22″][et_pb_row _builder_version=”3.22″ background_size=”initial” background_position=”top_left” background_repeat=”repeat”][et_pb_column type=”1_4″ _builder_version=”3.0.47″][et_pb_text _builder_version=”3.22.1″ background_color=”#072c72″ border_color_all=”#3255c9″ text_orientation=”right” background_layout=”dark” custom_padding=”20px|15px|15px|” z_index_tablet=”500″]<\/p>\n

2024 <\/em><\/p>\n

Strategic Projects<\/em><\/p>\n

@Chemistry<\/span><\/p>\n

[\/et_pb_text][et_pb_text _builder_version=”3.22.1″ text_orientation=”right” z_index_tablet=”500″]<\/p>\n

#simulated bodies<\/p>\n

#amortised inferences<\/p>\n

#inductive bias<\/p>\n

#fetus<\/p>\n

#MRI<\/p>\n<\/p>\n

[\/et_pb_text][\/et_pb_column][et_pb_column type=”3_4″ _builder_version=”3.0.47″][et_pb_text _builder_version=”3.22.1″ z_index_tablet=”500″]<\/p>\n

Project Summary<\/strong><\/h3>\n

Air pollution is the biggest environmental risk to health, which highlights the need for effective estimation of pollutant emissions. Improvements in satellite remote sensing technologies appear to be a game changer to better estimate pollutant emissions, yet it remains challenging to exploit the large amount of remote sensing data at high resolution. In this project, we draw on advanced deep learning techniques to address this challenge. With a focus on estimating nitrogen oxides (NOx) emissions from nitrogen dioxide (NO2) columns over East China. We first develop a data-driven deep neural network to emulate the state-of-the-art CHIMERE chemistry-transport model. We then embed it in a Physics-Informed Neural Network (PINN) for NOx emission estimation. We also utilize unsupervised domain adaptation techniques to handle the domain shift between varying data sources. The developed deep model will be applied to real NO2 data for 2019 to evaluate the impact of emission mitigation and for 2020 to evaluate the impact of COVID-19 on Chinese emissions.<\/p>\n

\u00a0<\/h5>\n

Ga\u00eblle Dufour
<\/span><\/strong>gaelle.dufour@lisa.ipsl.fr<\/h3>\n