the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 16 Feb 2024
Uncertainty of simulated brightness temperature due to sensitivity to atmospheric gas spectroscopic parameters
Abstract. Atmospheric radiative transfer models are extensively used in Earth observation to simulate radiative processes occurring in the atmosphere and to provide both upwelling and downwelling synthetic brightness temperatures for ground-based, airborne, and satellite radiometric sensors. For a meaningful comparison between simulated and observed radiances, it is crucial to characterise the uncertainty of such models. The purpose of this work is to quantify the uncertainty in radiative transfer models due to uncertainty in the associated spectroscopic parameters, and to compute simulated brightness temperature uncertainties for millimeter- and submillimeter-wave channels of downward-looking satellite radiometric sensors (MWI, ICI, MWS and ATMS) as well as upward looking airborne radiometers (ISMAR and MARSS). The approach adopted here is firstly to study the sensitivity of brightness temperature calculations to each spectroscopic parameter separately, then to identify the dominant parameters and investigate their uncertainty covariance, and finally to compute the total brightness temperature uncertainty due to the full uncertainty covariance matrix for the identified set of relevant spectroscopic parameters. The approach is applied to a recent version of the Millimiter-Wave propagation model, taking into account water vapor, oxygen, and ozone spectroscopic parameters, though it is general and can be applied to any radiative transfer code. A set of 135 spectroscopic parameters were identified as dominant for the uncertainty of simulated brightness temperatures (26 for water vapor, 109 for oxygen, none for ozone). The uncertainty of simulated brightness temperatures is computed for six climatology conditions (ranging from sub-Arctic winter to Tropical) and all instrument channels. Uncertainty is found to be up to few kelvin [K] in the millimeter-wave range, whereas it is considerably lower in the submillimeter-wave range (less than 1 K).
-
Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
-
Preprint
(1722 KB)
-
Supplement
(26 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(1722 KB) - Metadata XML
-
Supplement
(26 KB) - BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2023-3160', Anonymous Referee #1, 18 Mar 2024
This is a well-written paper describing a thorough study evaluating the uncertainties in space- and air-borne sensor channel radiance simulations due to uncertainties in the underlying spectroscopic parameters used in those simulations. This is of relevance to anyone working with simulated radiances for microwave sensors in a variety of applications. My comments below are minor, and after the authors consider them, I recommend that the paper be published.
Specific comments
Where the tables 1-3 indicate a range of estimated uncertainty for a parameter, what uncertainty value was used in the sensitivity analysis? (This should be described in the text).
Eqn 4: should the limit be as "\nu_0 -> 0" rather than "\nu -> 0"?
Lines 237-239. This sentence is confusing to me. I read it as saying that most channel passbands are at least 100 MHz from line centres (OK so far), and then it goes on to say that this is true for ICI channels at 325.15+/-1.5 GHz (sidebands 1600 MHz wide) and 664+/-4.2 GHz (sidebands 5000 MHz wide). In these examples, the sideband widths are larger than the offset from the central frequency meaning they overlap around the central frequency, so surely these bands do include the spectral line centre?
Typographical:
Line 21: "RTM represents" -> "RTM represent" (consistent use of RTM == radiative transfer models plural).
Line 71: "from 183 GHz and 664 GHz" -> "from 183 GHz to 664 GHz".
Inconsistent use of "vapour" and "vapor" throughout the text.
Citation: https://doi.org/10.5194/egusphere-2023-3160-RC1 -
AC1: 'Reply on RC1', Donatello Gallucci, 21 Mar 2024
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2023-3160/egusphere-2023-3160-AC1-supplement.pdf
-
AC1: 'Reply on RC1', Donatello Gallucci, 21 Mar 2024
-
RC2: 'Comment on egusphere-2023-3160', Anonymous Referee #2, 24 Mar 2024
The purpose of this work is to quantify the uncertainty in the radiative transfer models used to emulate microwave-to-submillimeter radiance observations from a range of sensors to the uncertainties in the spectroscopic parameters. The authors looked at all of the spectroscopic uncertainties, and after a sensitivity study, focused in on the more important values. It is very similar to a ground-breaking paper published by Cimini et al. (2018), but this work extends the first paper by looking at different viewing geometry, a different (and heavily utilized) absorption model, and a much larger frequency range.
Generally speaking, this paper is very strong, is well written, and makes a strong addition to the original paper. I have only a few concerns that should be addressed, and thus recommend minor revisions.
- The title should be expanded to indicate that this study is focused on the microwave to submillimeter. (A trivial change)
- The authors clearly understand the importance of accounting for the covariance in the uncertainties between different parameters. The easiest example to state is that the uncertainties in Cf and Cs (the continuum coefficients) are anti-correlated; this has been well known for years and indeed was discussed. However, this leads to two concerns:
- While the paper alludes to accounting for these covariances, there is no explicit statement on how this was determined and what those covariances were assumed to be in this analysis. This could be addressed with two additional tables (one for water vapor, one for oxygen) that provide correlation values between parameters (and when connected with table 1 or table 2 could be converted by the reader (like me) into covariances). I understand that there could be (and probably is) some spectral variability between the correlation of any two parameters, but still even having mean values of the correlations would be useful.
- One of the most fascinating plots in the Cimini et al. 2018 paper was Figure 9, which showed the covariance in the resulting Tb calculation from the spectroscopic parameter uncertainties. There is no information like this in this current paper, and it is essential before it be accepted for publication.
END
Citation: https://doi.org/10.5194/egusphere-2023-3160-RC2 -
AC2: 'Reply on RC2', Donatello Gallucci, 23 Apr 2024
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2023-3160/egusphere-2023-3160-AC2-supplement.pdf
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2023-3160', Anonymous Referee #1, 18 Mar 2024
This is a well-written paper describing a thorough study evaluating the uncertainties in space- and air-borne sensor channel radiance simulations due to uncertainties in the underlying spectroscopic parameters used in those simulations. This is of relevance to anyone working with simulated radiances for microwave sensors in a variety of applications. My comments below are minor, and after the authors consider them, I recommend that the paper be published.
Specific comments
Where the tables 1-3 indicate a range of estimated uncertainty for a parameter, what uncertainty value was used in the sensitivity analysis? (This should be described in the text).
Eqn 4: should the limit be as "\nu_0 -> 0" rather than "\nu -> 0"?
Lines 237-239. This sentence is confusing to me. I read it as saying that most channel passbands are at least 100 MHz from line centres (OK so far), and then it goes on to say that this is true for ICI channels at 325.15+/-1.5 GHz (sidebands 1600 MHz wide) and 664+/-4.2 GHz (sidebands 5000 MHz wide). In these examples, the sideband widths are larger than the offset from the central frequency meaning they overlap around the central frequency, so surely these bands do include the spectral line centre?
Typographical:
Line 21: "RTM represents" -> "RTM represent" (consistent use of RTM == radiative transfer models plural).
Line 71: "from 183 GHz and 664 GHz" -> "from 183 GHz to 664 GHz".
Inconsistent use of "vapour" and "vapor" throughout the text.
Citation: https://doi.org/10.5194/egusphere-2023-3160-RC1 -
AC1: 'Reply on RC1', Donatello Gallucci, 21 Mar 2024
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2023-3160/egusphere-2023-3160-AC1-supplement.pdf
-
AC1: 'Reply on RC1', Donatello Gallucci, 21 Mar 2024
-
RC2: 'Comment on egusphere-2023-3160', Anonymous Referee #2, 24 Mar 2024
The purpose of this work is to quantify the uncertainty in the radiative transfer models used to emulate microwave-to-submillimeter radiance observations from a range of sensors to the uncertainties in the spectroscopic parameters. The authors looked at all of the spectroscopic uncertainties, and after a sensitivity study, focused in on the more important values. It is very similar to a ground-breaking paper published by Cimini et al. (2018), but this work extends the first paper by looking at different viewing geometry, a different (and heavily utilized) absorption model, and a much larger frequency range.
Generally speaking, this paper is very strong, is well written, and makes a strong addition to the original paper. I have only a few concerns that should be addressed, and thus recommend minor revisions.
- The title should be expanded to indicate that this study is focused on the microwave to submillimeter. (A trivial change)
- The authors clearly understand the importance of accounting for the covariance in the uncertainties between different parameters. The easiest example to state is that the uncertainties in Cf and Cs (the continuum coefficients) are anti-correlated; this has been well known for years and indeed was discussed. However, this leads to two concerns:
- While the paper alludes to accounting for these covariances, there is no explicit statement on how this was determined and what those covariances were assumed to be in this analysis. This could be addressed with two additional tables (one for water vapor, one for oxygen) that provide correlation values between parameters (and when connected with table 1 or table 2 could be converted by the reader (like me) into covariances). I understand that there could be (and probably is) some spectral variability between the correlation of any two parameters, but still even having mean values of the correlations would be useful.
- One of the most fascinating plots in the Cimini et al. 2018 paper was Figure 9, which showed the covariance in the resulting Tb calculation from the spectroscopic parameter uncertainties. There is no information like this in this current paper, and it is essential before it be accepted for publication.
END
Citation: https://doi.org/10.5194/egusphere-2023-3160-RC2 -
AC2: 'Reply on RC2', Donatello Gallucci, 23 Apr 2024
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2023-3160/egusphere-2023-3160-AC2-supplement.pdf
Peer review completion
Journal article(s) based on this preprint
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 434 | 97 | 27 | 558 | 28 | 12 | 13 |
- HTML: 434
- PDF: 97
- XML: 27
- Total: 558
- Supplement: 28
- BibTeX: 12
- EndNote: 13
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Donatello Gallucci
Emma Turner
Stuart Fox
Philip W. Rosenkranz
Mikhail Y. Tretyakov
Vinia Mattioli
Salvatore Larosa
Filomena Romano
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(1722 KB) - Metadata XML
-
Supplement
(26 KB) - BibTeX
- EndNote
- Final revised paper