Spaceborne EO and a Combination of Inverse and Forward Modelling for Monitoring Lava Flow Advance

Round 1

Reviewer 1 Report

In this paper, authors investigate the relation between emissivity and temperature by performing laboratory measurements. The impact of a two components emissivity approach on estimates of radiant flux, and on numerical modelling, is then assessed. Although the manuscript is interesting, easy to read and fairly well structured, it requires revisions. Indeed, the abstract may be shortened, whereas both introduction and discussion section need to be extended. In the introduction, authors discuss advantages offered by satellite remote sensing in studying and monitoring thermal volcanic activity. However, there is no mention of satellite sensors currently used for this purpose, including the most recent ones (e.g. VIIRS and MSI) and relative studies. In addition, performed analyses and achieved results have to be discussed in more detail by the authors. Finally, some figures do not favor the interpretation of results. In the following, authors can find my specific comments and suggestions.

Specific comments:

Line 18-22: probably, you can remove this content.

Line 22: .. in radiant flux estimations and..

Line 36: Landsat 8 OLI data ?

Line 38-39: please, provide the full name of MODIS and SEVIRI.

Line 53: Do you refer to satellite remote sensing ? This should be specified here.

Line 58: please, specify the sensors. 

Line 59: ..during effusive eruptions

Line 62-63: this sentence shoud be moved before that at the line 57.

Line 64: the sentence should be modified as: “Sensors having channels in the Thermal Infrared bands (e.g. …)

Line 64: “..to detect volcanic thermal anomalies and have proven…”

Line 64-66: recent studies investigated thermal volcanic activity using also instruments as the MSI onboard Sentinel 2 satellites. This should be mentioned by the authors.

Line 66: volcano monitoring, this concept is repeated twice in the sentence.

Line 67: in my opinion, this sentence is too generic and should be reformulated.

Line 71: Landsat 8 OLI data? Please, specify.

Line 82: affect estimates of total and max radiant fluxes.

Line 83: “often neglected …” this concept is the same at lines 77-78;

Line 90: ..,we compared the results from spaceborne data with models…

Line 96: Mt Etna (Italy)

Line 97: “.. and styles..” provide one or more reference here;

Line 103-104: here you should refer to Figure 1. Please, check through the entire manuscript.

Line 107-108: see previous comment.

Line 112: In Figure 1, Mt. Etna geographic location needs to be better indicated by the authors (panel at the bottom right side of the figure is too small).

Line 121: section 2.1 is too short. I suggest authors to move Table 1 in this section, providing a brief description of its content.

Line 143: please, provided the full name of OLI and TIRS sensors. Main fetaures of these sensors should be also mentioned by the authors. 

Line 149: ..that are Line 150: and thermal volcanic activity.

Line 162: here we used the approach described in [27].

Line 177: who distributed the Landsat 8 OLI data used in this work? This should be specified here or in the acknowledgments.

Line 187: this formula may be removed.

Line 212: check the equation numbering. This should be the equation (5).

Line 216: see previous comment.

Line 222: which formulation was used to retrieve the radiant flux? The simplified version or the formulation in equation (5)? How was computed the atmospheric transmissivity and defined the shape of radiating surface?

Line 223: B6 and B7 are not defined in the text. 

Line 247: formula in equation (6) may be used to estimate the radiant flux from hot surfaces within a certain range of temperatures. This should be specified here.

Line 247-258: please, report all the terms in equation (6) in Italic.

Line 297: TIR wavelengths.

Line 232: MODIS and SEVIRI (for which authors should provide the full name) have to be mentioned before in the paper (i.e. introduction; see general comments). Moreover, authors should briefly describe their features (they could summarized in a Table).

Line 269: please provide one or more reference for Pléiades.

Line 272: please provide at least one reference here.

Line 298: ..decreases with..

Line 302: Authors should briefly discuss differences in emissivity values at the two analysed TIR wavelengths.

Line 308: Landsat 8 OLI data? please, specify both in the text and figure caption.

Line 311: this sentence is confusing and needs to be rewritten. “SWIR” has already been defined at line 69.

Line 313: temperature range mentioned in section 2.2.1 does not fit with the temperature of an active lava flow. Hence, emissivity values for crusted and molten components may be different from those considered here. This should be discussed in section .

Line 314: you could refer to hotspot pixels in this sentence.

Line 316: To estimate…

Line 318: How did you detect hotspot pixels before using the dual band approach? Did you apply a fixed threshold test to B7 radiance? This should be clarified in the text.

Line 322-326. Please, improve the figure quality.

Line 327: Figure 5. a)...

Line 328: how did you select the thresholds? What means BKGR 1.5? The uncertainty in detecting hotspot pixels on Landsat 8 OLI data and the relative impact on radiant flux estimations should be discussed by the authors in section 4. 

Line 331: Estimation of Total Radiat Flux from Landsat 8 OLI data…

Line 331: In table 2, columns “Payload” and “Pixel Area” may be removed, whereas components should be indicated in the two-component emissivity column.

Line 337: provide one or more references here.

Line 339: see previous comment.

Line 341: Which is the value of the radiant flux from infrared SEVIRI data at the time of Landsat 8 OLI observations?

Line 348, Figure 6: Please, improve the image resolution and add a label to help reader in better interpreting results.

Line 352: you mentioned MODIS and SEVIRI. Hence, you should refer to OLI data.

Line 355-356: ..suggests a low ratio..

Line 357:  please specify day, month and year on x-axis.

Line 358: see comment at line 348.

Line 358: see comment at line 352; I suggest to change the sentence as: .. recorded between.. and retrieved from..

Line 360: check the paragraph numbering

Line 391: there is no Table 3 in the figure.

Line 399-402: there are too brackets in this sentence.

Line 410-412: this sentence should be moved in the conclusions.

Line 419-422: this relevant information should be mentioned also in the abstract.

Author Response

c

To: Reviewer #1

Manuscript ID. remote sensing-645551 “Spaceborne EO and a combination of inverse and forward modelling for monitoring lava flow advance” by Nikola Rogic, Annalisa Cappello, Gaetana Ganci, Alessandro Maturilli, Hazel Rymer, Stephen Blake and Fabrizio Ferrucci.

Dear Reviewer,

We accepted and inserted all the corrections you suggested in the revised version. We very much appreciate your constructive criticism, which helped us improve the manuscript.

At point-by-point response to the concerns you raised by the reviewers is provided below. An annotated copy of the manuscript, where all the changes are track-coloured, has now been uploaded.

We report the changes made on the manuscript following the referee’s suggestions:

Your specific comments:

Line 18-22: probably, you can remove this content

Content removed

Line 22: ..in radiant flux estimations and..

Lines 18-22 have been removed

Line 36: Landsat 8 OLI data?

OLI specified

Line 38-39: please, provide the full name of MODIS and SEVIRI.

Full names provided

Line 53: Do you refer to satellite remote sensing? This should be specified here.

Specified

Line 58: please, specify the sensors.

We added the different spaceborne sensors used in this study MODIS, SEVIRI and OLI

Line 59: ..during effusive eruptions

Changed

Line 62-63: this sentence should be moved before that at the line 57.

Moved

Line 64: the sentence should be modified as: “Sensors having channels in the Thermal Infrared bands (e.g. …)

Suggestion accepted and added

Line 64: “..to detect volcanic thermal anomalies and have proven…”

Changed

Line 64-66: recent studies investigated thermal volcanic activity using also instruments as the MSI onboard Sentinel 2 satellites. This should be mentioned by the authors.

MSI onboard Sentinel 2 mentioned and referenced

Line 66: volcano monitoring, this concept is repeated twice in the sentence.

Removed repeated

Line 67: in my opinion, this sentence is too generic and should be reformulated. Reformulated

Line 71: Landsat 8 OLI data? Please, specify.

OLI specified

Line 82: affect estimates of total and max radiant fluxes.

Corrected

Line 83: “often neglected …” this concept is the same at lines 77-78;

Lines 77-78 have been removed

Line 90: ..,we compared the results from spaceborne data with models…

Corrected

Line 96: Mt Etna (Italy)

Changed

Line 97: “.. and styles..” provide one or more reference here

References provided

Line 103-104: here you should refer to Figure 1. Please, check through the entire manuscript.

Changed and checked

Line 107-108: see previous comment.

Line 112: In Figure 1, Mt. Etna geographic location needs to be better indicated by the authors (panel at the bottom right side of the figure is too small).

We improved Figure 1 to indicate Mt. Etna geographic location – panel changed, and we included scale and coordinates.

Line 121: section 2.1 is too short. I suggest authors to move Table 1 in this section, providing a brief description of its content.

Suggestion accepted: Table 1 moved, and a brief description and relevant references relating to samples’ composition added.

Line 143: please, provided the full name of OLI and TIRS sensors. Main features of these sensors should be also mentioned by the authors

Full names provided and sensors spectral range mentioned

Line 149: ..that are Line 150: and thermal volcanic activity.

Corrected

Line 162: here we used the approach described in [27].

Corrected

Line 177: who distributed the Landsat 8 OLI data used in this work? This should be specified here or in the acknowledgments.

GloVis specified and reference added

Line 187: this formula may be removed.

Formula removed

Line 212: check the equation numbering. This should be the equation (5).

Kept simplified equation (5) and removed expanded version.

Line 216: see previous comment.

Line 222: which formulation was used to retrieve the radiant flux? The simplified version or the formulation in equation (5)? How was computed the atmospheric transmissivity and defined the shape of radiating surface?

The simplified version was used to retrieve radiant flux, so the other version of Equation 5 was removed. Atmospheric transmissivity for the specific location (Etna) and date was calculated using Barsi et al. 2005 approach. 

Line 223: B6 and B7 are not defined in the text. 

Bands defined – band wavelengths were added.

Line 247: formula in equation (6) may be used to estimate the radiant flux from hot surfaces within a certain range of temperatures.

Clarified: The computation of the radiant heat flux in equation (6) is based on the approximation of the Planck Law in MIR to the 4th order power law, and this could introduce errors larger than 30% outside the range of temperatures ~600 – 1500 K, where that approximation holds (Wooster, M., Zhukov, B., & Oertel, D. 2003. Fire radiative energy release for quantitative study of biomass burning: derivation from the BIRD experimental satellite and comparison to MODIS fire products. Remote Sensing of Environment, 86, 83–107.)

Line 247-258: please, report all the terms in equation (6) in Italic.

Changed

Line 297: TIR wavelengths.

Changed

Line 232: MODIS and SEVIRI (for which authors should provide the full name) have to be mentioned before in the paper (i.e. introduction; see general comments). Moreover, authors should briefly describe their features (they could summarized in a Table).

We modified the Introduction in order to mention MODIS and SEVIRI. Moreover, we added the main features of these sensor in chapter 2.3.1.

Line 269: please provide one or more reference for Pléiades.

Reference added

Line 272: please provide at least one reference here.

Reference added

Line 298: ..decreases with..

Changed

Line 302: Authors should briefly discuss differences in emissivity values at the two analysed TIR wavelengths

Brief discussion added

Line 308: Landsat 8 OLI data? please, specify both in the text and figure caption.

OLI has been specified

Line 311: this sentence is confusing and needs to be rewritten. “SWIR” has already been defined at line 69.

Rewritten

Line 313: temperature range mentioned in section 2.2.1 does not fit with the temperature of an active lava flow. Hence, emissivity values for crusted and molten components may be different from those considered here. This should be discussed in section

Clarified in section 4: In this study, our experiments for emissivity measurements include a temperature range of 400-900 K, which can be related to cooling and crusted flows. For the molten portion of the flow, we used emissivity values retrieved from literature (Ramsey et al., 2019). Indeed, as stated in the Conclusions, we intend to extend our laboratory measurements to include higher, realistic volcanic eruptive temperatures (up to ~1350 K for Etnean lavas).

Line 314: you could refer to hotspot pixels in this sentence.

Changed to hotspot pixels

Line 316: To estimate…

Corrected

Line 318: How did you detect hotspot pixels before using the dual band approach? Did you apply a fixed threshold test to B7 radiance? This should be clarified in the text.

Clarified

Line 322-326. Please, improve the figure quality.

Figures 5. (a) and (b) changed - improved quality

Line 327: Figure 5. a)...

Corrected

Line 328: how did you select the thresholds? What means BKGR 1.5? The uncertainty in detecting hotspot pixels on Landsat 8 OLI data and the relative impact on radiant flux estimations should be discussed by the authors in section 4

Background Radiance (BKGR 1.5) specified and the uncertainty has been clarified

Line 331: Estimation of Total Radiant Flux from Landsat 8 OLI data…

Changed

Line 331: In table 2, columns “Payload” and “Pixel Area” may be removed, whereas components should be indicated in the two-component emissivity column.

Suggested columns removed and components indicated - component 1* (assumed crust) & component 2** (assumed melt) below the Table 2.

Line 337: provide one or more references here.

SEVIRI, enables a near-continuous monitoring of volcanic thermal activity (Referenced)

Line 339: see previous comment.

MODIS, permits detection of less intense thermal anomalies (Referenced)

Line 341: Which is the value of the radiant flux from infrared SEVIRI data at the time of Landsat 8 OLI observations?

For dates/times in Table 2

1)    Landsat 8 (OLI) 18/03/2017 at 09:41, radiant flux 3.12 GW to 3.68 GW

SEVIRI 18/03/2017 at 09:41, radiant flux 2.75 GW to 6.88 GW

2)    Landsat 8 (OLI) 27/03/2017 at 09:35, radiant flux 1.68 GW to 1.97 GW

SEVIRI 27/03/2017 at 08:57, radiant flux 1.43 GW to 3.58 GW

For these reasons we stated that OLI data fits ‘lower boundaries’ in section 4.

Line 348, Figure 6: Please, improve the image resolution and add a label to help reader in better interpreting results.

We improved figure 6 by adding the legend and DATE (dd-mmm-yy) on x-axis.

Line 352: you mentioned MODIS and SEVIRI. Hence, you should refer to OLI data.

OLI mentioned

Line 355-356: ..suggests a low ratio..

Changed

Line 357:  please specify day, month and year on x-axis

We specified DATE (dd-mmm-yy) on x-axis

Line 358: see comment at line 348

Figure 7 changed

Line 358: see comment at line 352; I suggest to change the sentence as: .. recorded between.. and retrieved from..

Suggestion accepted – text changed. The peak of activity is not recorded between 17 and 18 March. It’s recorded on 17 March by MODIS and on 18 March by SEVIRI.

Line 360: check the paragraph numbering

Checked and changed to 3.2.4.

Line 391: there is no Table 3 in the figure.

Changed to Table 2

Line 399-402: there are two brackets in this sentence.

Brackets removed

Line 410-412: this sentence should be moved in the conclusions.

Sentence moved to conclusions

Line 419-422: this relevant information should be mentioned also in the abstract.

This information has now been mentioned in the abstract

Reviewer 2 Report

Assessment  of the emissivity role in the determination of radiant fluxes obtained from space-borne thermal sensors is crucial research component. With the present study authors compare the results obtained from  the space-borne thermal data and compare them with models using 3 different approaches and using as test site the Etna volcano among different effusive events the episode started from the old “saddle” between the South‐East Crater (SEC) 103 and the New Southeast Crater (NSEC) on the morning of 15 March 2017 has been chosen. Finally the present paper increase our  knowledge about the behavior of emissivity at molten rock temperatures.

Comments

line 142 High-spatial resolution data

I am not sure that Landsat  8 is characterized as high spatial resolution. Definitely compared to the SEVIRI data they could be characterized as such.

lines  148-229 This part constitutes a generic introductory section on the temperature behavior of objects and thermal remote sensing could be part of introduction with less lines.

line 268 Can you specify more about the processing a tri-stereo Pleiades imagery that you used?

lines 311-314 Following Planck’s Law, the high-temperature thermal anomalies from our lava flow are best  observed in Short-Wavelength Infra-Red (SWIR), as this region of the electromagnetic spectrum is  more sensitive to the temperatures found in an active lava flow (i.e. ~900 to 1300 K) than the TIR region. Please give some references.

Question

How the different wavelengths (low and high temporal resolution data)  affect measured temperatures?

How the morphology, and specifically DEM-derived spatial models like slope and aspect affect the Land surface temperature (LST) and land surface emissivity (LSE) and generally the results of the experiments in this paper?

Author Response

To: Reviewer #2

Manuscript ID. remote sensing-645551 “Spaceborne EO and a combination of inverse and forward modelling for monitoring lava flow advance” by Nikola Rogic, Annalisa Cappello, Gaetana Ganci, Alessandro Maturilli, Hazel Rymer, Stephen Blake and Fabrizio Ferrucci.

Dear Reviewer,

We accepted and inserted all the corrections you suggested in the revised version. We very much appreciate your constructive criticism, which helped us improve the manuscript.

At point-by-point response to the concerns you raised by the reviewers is provided below. An annotated copy of the manuscript, where all the changes are track-coloured, has now been uploaded.

We report the changes made on the manuscript following the referee’s suggestions:

line 142 High-spatial resolution data

I am not sure that Landsat 8 is characterized as high spatial resolution. Definitely compared to the SEVIRI data they could be characterized as such.

Landsat 8 OLI data (30 m) in comparison to MODIS (1000 m) and SEVIRI (3000 m) can be considered of a high-spatial resolution for the purpose of this study.

lines 148-229 This part constitutes a generic introductory section on the temperature behavior of objects and thermal remote sensing could be part of introduction with less lines

Parts moved to introduction and shortened

line 268 Can you specify more about the processing a tri-stereo Pleiades imagery that you used?

We clarified that the 3D processing of the tri-stereo Pléiades imagery was performed using the free and open source MicMac photogrammetric library (http://micmac.ensg.eu) developed by the French Institut Géographique National.

lines 311-314 Following Planck’s Law, the high-temperature thermal anomalies from our lava flow are best observed in Short-Wavelength Infra-Red (SWIR), as this region of the electromagnetic spectrum is more sensitive to the temperatures found in an active lava flow (i.e. ~900 to 1300 K) than the TIR region. Please give some references

Paragraph rewritten and referenced (Harris, 2013)

Question

How the different wavelengths (low and high temporal resolution data) affect measured temperatures?

Added to section 3.2.1 relating to low, moderate and high temporal resolution data: According to the Wien's displacement law, the wavelength of maximum radiance shifts to shorter wavelengths with increasing temperature. The temperature of volcanic features varies widely, ranging between 600 and 1500 K, and thus can be observed in TIR, MIR and SWIR portions of the spectrum. SWIR radiation is mainly emitted in detectable quantities from very hot surfaces (e.g., molten lava), while MIR and TIR also detect cooler volcanic surfaces.

How the morphology, and specifically DEM-derived spatial models like slope and aspect affect the Land surface temperature (LST) and land surface emissivity (LSE) and generally the results of the experiments in this paper?

Added to section 4: The terrain can attenuate the observed spectral radiance, and hence the LST, because it can act as an obstacle in function of the mutual angle between the sensor plane of the satellite instrument and the normal direction to each cell of the DEM. LSE is an indicator of material composition and can vary with viewing angle and surface roughness. In our case study, i.e. the 2017 Etna eruption, the median slope is about 6° and there are no shadows due to the topography, so the terrain contribution can be neglected.

Round 2

Reviewer 1 Report

The paper has been significantly improved. Authors have modified both manuscript and figures taking into account all my suggestions.

Therefore I have only a few minor comments here listed:

Line 77: please change the sentence as MODIS onboard Terra and Aqua satellites, SEVIRI onboard the Meteosat Second Generation (MSG) geostationary platform etc. Line 87: you could mention the Stefan Boltzmann law without reporting any equation in the introduction.

Line 97: I suggest to change the sentence as: “Sensors having channels in the Mid-Infrared (MIR) and Thermal Infrared (TIR) bands (e.g. MODIS) have been…. “

Line 100: “On the other hand, sensors such as the Multi-Spectral Imager (MSI) aboard Sentinel-2 satellites and OLI having channels in the Short-Wave Infrared (SWIR) region, and providing data at medium-high spatial resolution, enable a better identification of high-temperature thermal anomalies (e.g. lava flows)“. You could refer here to section 3.2.1

Line 395: ambient..

Line 436: check the sentence

Author Response

To: Reviewer 1

Manuscript ID. remote sensing-645551 “Spaceborne EO and a combination of inverse and forward modelling for monitoring lava flow advance” by Nikola Rogic, Annalisa Cappello, Gaetana Ganci, Alessandro Maturilli, Hazel Rymer, Stephen Blake and Fabrizio Ferrucci.

Dear reviewer,

We accepted and inserted, in the revised version, all the corrections you suggested. We very much appreciate your constructive criticism, which helped us improve the manuscript.

Best wishes,

Nikola Rogic

We report the changes made on the manuscript following the referee’s suggestions:

REVIEWER #1

Comments and Suggestions for Authors

Line 77: please change the sentence as MODIS onboard Terra and Aqua satellites, SEVIRI onboard the Meteosat Second Generation (MSG) geostationary platform etc. Line 87: you could mention the Stefan Boltzmann law without reporting any equation in the introduction.

Sentence changed in line 77 and Stefan Boltzmann equation 1 has been removed. Equations are now renumbered 1-5, instead of 1-6.

Line 97: I suggest to change the sentence as: “Sensors having channels in the Mid-Infrared (MIR) and Thermal Infrared (TIR) bands (e.g. MODIS) have been…. “

Suggestion accepted and changed

Line 100: “On the other hand, sensors such as the Multi-Spectral Imager (MSI) aboard Sentinel-2 satellites and OLI having channels in the Short-Wave Infrared (SWIR) region, and providing data at medium-high spatial resolution, enable a better identification of high-temperature thermal anomalies (e.g. lava flows)“. You could refer here to section 3.2.1

Changed

Line 395: ambient..

Changed

Line 436: check the sentence

Checked