A neural network radiative transfer model approach applied to the Tropospheric Monitoring Instrument aerosol height algorithm

Abstract. To retrieve aerosol properties from satellite measurements of the oxygen A-band in the near-infrared, a line-by-line radiative transfer model implementation requires a large number of calculations. These calculations severely restrict a retrieval algorithm's operational capability as it can take several minutes to retrieve the aerosol layer height for a single ground pixel. This paper proposes a forward modelling approach using artificial neural networks to speed up the retrieval algorithm. The forward model outputs are trained into a set of neural network models to completely replace line-by-line calculations in the operational processor. Results comparing the forward model to the neural network alternative show an encouraging outcome with good agreement between the two when they are applied to retrieval scenarios using both synthetic and real measured spectra from TROPOMI (TROPOspheric Monitoring Instrument) on board the European Space Agency (ESA) Sentinel-5 Precursor mission. With an enhancement of the computational speed by 3 orders of magnitude, TROPOMI's operational aerosol layer height processor is now able to retrieve aerosol layer heights well within operational capacity.

[1]  John V. Martonchik,et al.  Retrieval of the optical depth and vertical distribution of particulate scatterers in the atmosphere using O 2 A- and B-band SCIAMACHY observations over Kanpur: a case study , 2012 .

[2]  Robert Frouin,et al.  Estimating the altitude of aerosol plumes over the ocean from reflectance ratio measurements in the O2 A-band. , 2009 .

[3]  Diego G. Loyola,et al.  Automatic cloud analysis from polar-orbiting satellites using neural network and data fusion techniques , 2004, IGARSS 2004. 2004 IEEE International Geoscience and Remote Sensing Symposium.

[4]  Eleni Marinou,et al.  An exploratory study on the aerosol height retrieval from OMI measurements of the 477 nmO 2 O 2 spectral band using a neural network approach , 2016 .

[5]  Abram F. J. Sanders,et al.  Error sources in the retrieval of aerosol information over bright surfaces from satellite measurements in the oxygen A-band , 2017 .

[6]  L. G. Tilstra,et al.  Interpretation of FRESCO cloud retrievals in case of absorbing aerosol events , 2011 .

[7]  J. Veefkind,et al.  Spatial distribution analysis of the OMI aerosol layer height: a pixel-by-pixel comparison to CALIOP observations , 2017 .

[8]  A weighted least squares approach to retrieve aerosol layer height over bright surfaces applied to GOME-2 measurements of the oxygen A band for forest fire cases over Europe , 2018, Atmospheric Measurement Techniques.

[9]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.

[10]  Yi Wang,et al.  Passive remote sensing of altitude and optical depth of dust plumes using the oxygen A and B bands: First results from EPIC/DSCOVR at Lagrange‐1 point , 2017, Geophysical research letters.

[11]  Adrian Doicu,et al.  The operational cloud retrieval algorithms from TROPOMI on board Sentinel-5 Precursor , 2017 .

[12]  L. C. Henyey,et al.  Diffuse radiation in the Galaxy , 1940 .

[13]  E. Kort,et al.  Constraining Aerosol Vertical Profile in the Boundary Layer Using Hyperspectral Measurements of Oxygen Absorption , 2018, Geophysical Research Letters.

[14]  D. Diner,et al.  Aerosol Layer Height over Water from O2 A-Band: Mono-Angle Hyperspectral and/or Bi-Spectral Multi-Angle Observations , 2017 .

[15]  Piet Stammes,et al.  Evaluation of the operational Aerosol Layer Height retrieval algorithm for Sentinel-5 Precursor: application to O 2 A band observations from GOME-2A , 2015 .

[16]  Piet Stammes,et al.  Surface reflectivity climatologies from UV to NIR determined from Earth observations by GOME‐2 and SCIAMACHY , 2017 .

[17]  Laurent Itti,et al.  Overcoming catastrophic forgetting problem by weight consolidation and long-term memory , 2018, ArXiv.

[18]  Abram F. J. Sanders,et al.  Retrieval of aerosol parameters from the oxygen A band in the presence of chlorophyll fluorescence , 2013 .

[19]  TROPOMI ATBD of the Aerosol Layer Height , 2019 .

[20]  V. Rozanov,et al.  Information content of the spectral measurements of the 0.76 μm O2 outgoing radiation with respect to the vertical aerosol optical properties , 1995 .

[21]  Clive D Rodgers,et al.  Inverse Methods for Atmospheric Sounding: Theory and Practice , 2000 .

[22]  Nicholas J. Nauslar,et al.  The 2017 North Bay and Southern California Fires: A Case Study , 2018, Fire.

[23]  J. Thepaut,et al.  The ERA‐Interim reanalysis: configuration and performance of the data assimilation system , 2011 .

[24]  M. V. Roozendael,et al.  FRESCO+: an improved O 2 A-band cloud retrieval algorithm for tropospheric trace gas retrievals , 2008 .

[25]  Alexander Vasilkov,et al.  Note on rotational-Raman scattering in the O 2 A- and B-bands , 2013 .

[26]  Otto P. Hasekamp,et al.  Efficient calculation of intensity and polarization spectra in vertically inhomogeneous scattering and absorbing atmospheres , 2008 .

[27]  Stanley P. Sander,et al.  A sensitivity study on the retrieval of aerosol vertical profiles using the oxygen A-band , 2015 .

[28]  C. Sioris,et al.  Impact of rotational Raman scattering in the O2A band , 2000 .

[29]  Robert Frouin,et al.  Retrieval of the aerosol vertical distribution from atmospheric radiance , 2008, Asia-Pacific Remote Sensing.

[30]  Juergen Fischer,et al.  Retrieving aerosol height from the oxygen A band : a fast forward operator and sensitivity study concerning spectral resolution , instrumental noise , and surface inhomogeneity , 2013 .

[31]  Marco Cervino,et al.  Aerosol extinction coefficient profile retrieval in the oxygen A-band considering multiple scattering atmosphere. Test case : SCIAMACHY nadir simulated measurements , 2006 .

[32]  Harlan G. Hughes.  Aerosol Extinction Coefficient Variations with Altitude at 3.75 , 1980 .

[33]  Thomas Trautmann,et al.  A linearized radiative transfer model for ozone profile retrieval using the analytical forward-adjoint perturbation theory approach , 2001 .

[34]  Christopher W. O'Dell,et al.  Aerosol information content analysis of multi-angle high spectral resolution measurements and its benefit for high accuracy greenhouse gas retrievals , 2012 .

[35]  Hartmut Bösch,et al.  Tropospheric aerosol profile information from high-resolution oxygen A-band measurements from space , 2015 .

[36]  Henk Eskes,et al.  TROPOMI on the ESA Sentinel-5 Precursor: A GMES mission for global observations of the atmospheric composition for climate, air quality and ozone layer applications , 2012 .

[37]  R. E. Wengert,et al.  A simple automatic derivative evaluation program , 1964, Commun. ACM.

[38]  M. Gabella,et al.  Retrieval of aerosol profile variations from reflected radiation in the oxygen absorption a band. , 1999, Applied optics.

[39]  J. Hovenier,et al.  The adding method for multiple scattering calculations of polarized light , 1987 .

[40]  Razvan Pascanu,et al.  Overcoming catastrophic forgetting in neural networks , 2016, Proceedings of the National Academy of Sciences.

[41]  Heikki Saari,et al.  The ozone monitoring instrument , 2006, IEEE Transactions on Geoscience and Remote Sensing.

[42]  Kelly Chance,et al.  An improved high-resolution solar reference spectrum for earth's atmosphere measurements in the ultraviolet, visible, and near infrared , 2010 .