Nitrous Oxide Profiling from Infrared Radiances (NOPIR): Algorithm Description, Application to 10 Years of IASI Observations and Quality Assessment

Round 1

Reviewer 1 Report

This paper retrieve N2O partial column (XN2O) from IASI spectra. The results are validated using TCCON measurements. The IASI XN2O values are compared with those from the TCCON network. The XN2O partical columns do not look to be easily explained but likely require improvements as the authors suggest. This is good progress for publication in Remote Sensing.

Author Response

We are grateful to the referee for his/her review of our work. Indeed, the reason for the difference between ground-based (NDACC, TCCON) and IASI is not yet fully explained, but that difference remains within the sum of the uncertainties of the compared data sets.

Reviewer 2 Report

Review of the study entitled “Nitrous Oxide Profiling from Infrared Radiances (NOPIR): algorithm description, application to 10 years of IASI observations and quality assessment” by Sophie Vandenbussche et al.

The study presents nitrous oxide (N2O) column retrievals from IASI satellite measurements using the nitrous oxide profiling from infrared radiances (NOPIR) algorithm. The N2O retrievals are evaluated against ground-based measurements at several ground-based NDACC and TCCON stations from the north to the south and small deviations are found. I am reading a well written and well-structured paper, and I am in favor to seeing it published after few minor comments and clarifications.

Lines 103-131: It would fit better as a subsection of section Methods.

Line 132-147: It would fit better in the conclusions.

Figure 10b and related lines 519-521: I cannot understand the correlation plot. There are no correlation values.

Figures 16-18: It is weird why such strong trends compared to GB data. I see that it is artificial as explained in the conclusions, but since trends of this gas are important, I was wondering if you have tested the GOME-2B data to see if match better with the GB measurements regarding the trends, and regarding the seasonal cycle shown in Figure 19 where clearly there is a positive bias of IASI data. If IASI MetOpB N2O data perform better than MetOpA data, that should be communicated to the community.

Good work toward progress to provide a global coverage of N2O using satellite measurements – congratulations.

Author Response

We are grateful to the referee for his/her review of our work and for his/her positive comments. Hereunder her/his comments are copied, and the answer is provided in blue italic.

Lines 103-131: It would fit better as a subsection of section Methods.

The lines were moved to the end of section 3 as a separate section 3.4 and a slight modification of the introductory sentences.

Line 132-147: It would fit better in the conclusions.

I agree with the referee, and it had already been discussed amongst co-authors. The only reason for these lines to be in the introduction is that the instructions for authors states that we should “highlight the main conclusions” in the introduction. Those lines were simply removed, as they were a summary of the discussion / conclusion section.

Figure 10b and related lines 519-521: I cannot understand the correlation plot. There are no correlation values.

Indeed, the correlation value was not provided because the goal of the figure (10a and 10b together) was to show the link between increasing satellite viewing zenith angle and retrieval number of degrees of freedom. The correlation coefficient itself could be impacted by other factors that have an impact on the DOF (like surface temperature, surface altitude) as can be seen on fig 10a: the DOF changes within the IASI swath (as a function of the viewing angle) are not the same at all locations. However, for the referee’s information: the correlation coefficient in fig 10b is 0.42. To avoid the same difficulty for future readers, I have removed the term “correlation” from the figure caption and the text referring to it.

Figures 16-18: It is weird why such strong trends compared to GB data. I see that it is artificial as explained in the conclusions, but since trends of this gas are important, I was wondering if you have tested the GOME-2B data to see if match better with the GB measurements regarding the trends, and regarding the seasonal cycle shown in Figure 19 where clearly there is a positive bias of IASI data. If IASI MetOpB N2O data perform better than MetOpA data, that should be communicated to the community.

It is not precisely a trend of the IASI bias, but a change in bias at (almost) each update in the IASI operational processing version. This shows the sensitivity of the N2O retrieval consistency to the temperature profile consistency. We have not tried IASI/MetOp-B data yet (we do not store it at our institute to limit the data storage amount) but I would absolutely not expect any difference in the inconsistencies: the IASI operational processor is updated at the same time for all available IASI instruments. So each “step” in the IASI-A time series would also be seen in the IASI-B and IASI-C time series – whenever they overlap. The 3 IASI instruments have been shown to be consistent with each other, as far as I know, with small differences linked to a different observation time (slightly different orbit). The trend analysis will have to be performed on a separate processing of the data, using IASI spectra and a consistent set of temperature profiles, either from a reprocessed consistent IASI version from EUMETSAT, or from model data, as ERA5.

As far as GOME-2 is concerned, I am not aware of a N2O product. I believe that the referee meant IASI-B in her/his comment. If not, please forgive me.