10.21413/IASI-FT_METOPA_ATP_L3_LATMOS-ULB
Bouillon, Marie
Marie
Bouillon
https://orcid.org/0000-0003-1675-0933
Atmospheres Laboratory Environments, Observations Spatiales
Sorbonne University
Versailles Saint-Quentin-en-Yvelines University
French National Centre for Scientific Research
IASI Atmospheric Temperature Profiles (ATP) from IASI-Metop-A
LATMOS/ULB
2021
Netcdf
FOS: Earth and related environmental sciences
FOS: Earth and related environmental sciences
Clerbaux, Cathy
Cathy
Clerbaux
https://orcid.org/0000-0003-0394-7200
Atmospheres Laboratory Environments, Observations Spatiales
Sorbonne University
Versailles Saint-Quentin-en-Yvelines University
French National Centre for Scientific Research
Bouillon, Marie
Marie
Bouillon
Atmospheres Laboratory Environments, Observations Spatiales
Sorbonne University
Versailles Saint-Quentin-en-Yvelines University
French National Centre for Scientific Research
George, Maya
Maya
George
https://orcid.org/0000-0001-8897-7964
Atmospheres Laboratory Environments, Observations Spatiales
Sorbonne University
Versailles Saint-Quentin-en-Yvelines University
French National Centre for Scientific Research
2021-06-01
2007-07-10/2017-12-31
2021
en
10.15770/eum_sec_clm_0014
The IASI atmospheric temperature profiles product (IASI ATP) is a daily 1°x1° global dataset of atmospheric temperatures derived from all IASI radiance observations. The temperatures profiles are given on 11 static pressure levels from 750 to 2 hPa. The method used to retrieve atmospheric temperatures from IASI radiances is detailed in Bouillon et al., 2021 (to be submitted very soon). It relies on an Artificial Neural Network (ANN) trained with 450 000 observations (IASI observations and NOAA-CO2 in input, ERA5 observations in output. The trained ANN was applied to reprocessed IASI L1C data, in order to produce a homogeneous atmospheric temperature record from 2008 to present. The dataset consists of daily mean NETCDF files of around 3 MB each (total 14 GB for the 2008-2017 period available to date). Each file includes the gridded daily atmospheric temperature profiles (day and night mixed together). Missing values are set at NaN. Size of the grids: 90 lines (from -89.5 to 89.5 °N), 180 columns (from -179.5 to 179.5 °E), 11 aisles (pressure levels: 2, 7, 10, 30, 20, 70, 100, 200, 400, 550 and 750 hPa). Unit: K.
The method is based on an artificial neural network (ANN) technique, trained to compute atmospheric temperatures from spectral radiances. For the training of the ANN, 450 000 IASI observations, and the corresponding ERA5 temperature profiles, were selected randomly over the period 2008-2020. The trained ANN was then applied to compute atmospheric temperatures from all IASI/Metop-A observations from 2007-07-10 to 2017-12-31 and all IASI/Metop-B observations from 2018-01-01 to 2020-12-31.
Product data level : "L3" (daily 1°x1° grid)
https://iasi.cnes.fr/
Publication: Bouillon, M., Safieddine, S., Whitburn, S., Clarisse, L., Aires, F., Pellet, V., Lezeaux, O., Scott, N., Doutriaux-Boucher, M. and Clerbaux, C. (2021). Time evolution of temperature profiles retrieved from 13 years of IASI data using an artificial neural network. (to be submitted soon in Atmospheric Measurement Techniques journal)