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

Nanda, Swadhin; de Graaf, Martin; Veefkind, J. Pepijn; ter Linden, Mark; Sneep, Maarten; de Haan, Johan; Levelt, Pieternel F.

doi:https://doi.org/10.5194/amt-12-6619-2019

Articles | Volume 12, issue 12

https://doi.org/10.5194/amt-12-6619-2019

© Author(s) 2019. This work is distributed under
the Creative Commons Attribution 4.0 License.

Special issue:

TROPOMI on Sentinel-5 Precursor: first year in operation (AMT/ACP...

https://doi.org/10.5194/amt-12-6619-2019

© Author(s) 2019. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 12, issue 12

Research article

|

13 Dec 2019

Research article |

| 13 Dec 2019

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

Swadhin Nanda, Martin de Graaf, J. Pepijn Veefkind, Mark ter Linden, Maarten Sneep, Johan de Haan, and Pieternel F. Levelt

Download

Final revised paper (published on 13 Dec 2019)
Preprint (discussion started on 08 May 2019)

Interactive discussion

Status: closed

AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment

- Printer-friendly version

- Supplement

RC1: 'Review of "A neural network radiative transfer model approach applied to TROPOMI’s aerosol height algorithm"', Anonymous Referee #1, 20 Jun 2019
- AC1: 'Response to amt-2019-143-RC1', Swadhin Nanda, 11 Jul 2019
RC2: 'Review', Anonymous Referee #2, 03 Jul 2019
- AC2: 'Response to amt-2019-143-RC2', Swadhin Nanda, 11 Jul 2019

Peer-review completion

AR: Author's response | RR: Referee report | ED: Editor decision

AR by Swadhin Nanda on behalf of the Authors (12 Jul 2019) Author's response Manuscript

ED: Referee Nomination & Report Request started (31 Jul 2019) by Jhoon Kim

RR by Anonymous Referee #3 (09 Sep 2019)

ED: Publish subject to minor revisions (review by editor) (21 Sep 2019) by Jhoon Kim

AR by Swadhin Nanda on behalf of the Authors (25 Sep 2019) Author's response Manuscript

ED: Publish subject to minor revisions (review by editor) (13 Oct 2019) by Jhoon Kim

This manuscript is much improved than before, but requires thorough check on sentences. Here are some examples:

Page 2, line 26 : O2-O2 O2-O2 (last '2' not in subscript)
Page2, line 28: MODIS aerosol optical depth at 550 nm product ==> MODIS aerosol optical depth at 550 nm.
Page 2, Line 30: 1) They demonstrated: here “they” should be matched to “their neural network models”, please amend. 2) OMI is here indicated such as “the Ozone Monitoring Instrument (OMI) on board the NASA Aura mission, it looks like that OMI is first mentioned here, but already mentioned many times such as line 26 and 29 in the same page.
Page 3. Some references are in wrong foramt, e.g. without year of publication: e.g., lines 3, 4, 6 (e.g., Loyola et al. instead of Loyola et al. 2018).
Page 4, line 14: n iterative estimate The nth iterative estimate?
In 2.1, any explanation for Sa?
Section 2.2, For simulating the measured reflectance, the high resolution reflectance () calculated from NN is directly convolved with instrument slit function ( ), but you convert reflectance to radiance (at high resolution and then perform the convolution for and , separately for finally getting I/F, which is indicated by “Io correction”. I am curious that there is significant impact on O2 absorption band with and without Io correction. I think that it could be very useful for readers to describe why you perform Io correction for your convolution process.
Page 5, line 2: aerosols, the surface and molecular species aerosols, surface, and molecular species ?
Page6, line 17: the ozone monitoring instrument (OMI) OMI
Page 8, line 33: a fixed g value the symbol g is not specified, it could be good to denote g in page 5, line 29 in section of 2.2. In addition, the symbol omega for single scattering albedo could be first denoted in section of 2.2.
Page 11, line 33: A histogram of these differences in plotted in Figure 6 is plotted
Page 13, line 1-2: “Absorbing aerosol index (AAI) should be placed in the first line rather than the second line.
Page 14, line 14: each of ~ were each of ~ was

Hide

AR by Swadhin Nanda on behalf of the Authors (20 Oct 2019) Author's response Manuscript

ED: Publish as is (03 Nov 2019) by Jhoon Kim

AR by Swadhin Nanda on behalf of the Authors (03 Nov 2019)

Short summary

This paper discusses a neural network forward model used by the operational aerosol layer height (ALH) retrieval algorithm for the TROPOspheric Monitoring Instrument (TROPOMI) on board the European Sentinel-5 Precursor satellite mission. This model replaces online radiative transfer calculations within the oxygen A-band, improving the speed of the algorithm by 3 orders of magnitude. With this advancement in the algorithm's speed, TROPOMI is set to deliver the ALH product operationally.