Role of Emissivity in Lava Flow ‘Distance-to-Run’ Estimates from Satellite-Based Volcano Monitoring
Jonathan Dehn
Round 1
Reviewer 1 Report
Review of the manuscript “The role of emissivity in lava flow ‘distance-to-run’ estimates from satellite-based volcano monitoring”, submitted to Remote Sensing by Rogic et al.
Rogic et al. present multi-stage experiment to derive emissivity, combining laboratory-based FTIR analyses, remote sensing data and numerical modeling. The manuscript is well written and scientifically sound, showing how uncertainties in emissivity affect “distance-to-run” estimates of a lava flow.
In my opinion, the main results of this work can be summarized as follow: 1) emissivities derived using FTIR analyses are in full agreement with those derived by ASTER GED; 2) the best fit between observed and modeled results is associated with the emissivity value derived by combined ASTER-FTIR analysis (0.93).
Moreover, these results convinced me that ASTER GED can be successfully used to have a best estimate of emissivities at volcanoes worldwide.
Unfortunately, I noticed that these results and conclusions are not clearly stated in the text. I then suggest to revise the manuscript considering these issues.
Specific comments:
222-231: Even though this paper is focused on emissivity uncertainties, authors should at least briefly discuss how uncertainties on other parameters involved in equations (ρ, CP, T0, Tf, CL and Φ) could affect effusion rate estimates. To derive eruptive parameters, are uncertainties in emissivity playing the primary role, respect to uncertainties associated with these other parameters?
243-244: I calculated a 20% emissivity change affects only for a 2% and 6% the temperature estimates of SWIR and TIR bands respectively. It seems from your results that emissivity does not play an important role in estimating temperature.
270-277: I calculated an Lmax increase of only 5% (300m/6400m) and not of 11% as stated. This paragraph is not clear. Please re-write it.
313-317: Why the best results in terms of area covered (1.94 km2 ) and average distribution (12.4 m) are obtained by the MAGFLOW simulation run with an emissivity of 0.8, while the best fit for length (6.4 km) is obtained by the simulation run with ε = 0.93? Please, discuss this issue.
338-339: It seems that a 20% emissivity change would affect distance-to-run estimation for only 5% (300m/6400m). This is a good result, which would rather suggest that emissivity uncertainties does not impact very much on hazard mitigation and planning.
p { margin-bottom: 0.1in; line-height: 120%; }
Author Response
Manuscript Ref. remotesensing-447306: “Role of emissivity in lava flow ‘distance-to-run’ estimates from satellite-based volcano monitoring” by Nikola Rogic, Annalisa Cappello, Fabrizio Ferrucci
Dear Reviewer # 1,
We accepted and inserted in the revised version all the corrections and you suggested. We very much appreciate your constructive criticism, which helped us improve the manuscript.
A point-by-point response to the concerns you raised is provided below. An annotated copy of the manuscript, where all the changes are track-coloured, has been uploaded.
We report the changes made on the manuscript following the referee’s suggestions:
Reviewer #1: 222-231: Even though this paper is focused on emissivity uncertainties, authors should at least briefly discuss how uncertainties on other parameters involved in equations (ρ, CP, T0, Tf, CL and Φ) could affect effusion rate estimates. To derive eruptive parameters, are uncertainties in emissivity playing the primary role, respect to uncertainties associated with these other parameters?
Authors:
· Both equations [2] and [3] have been adjusted and a paragraph was added to explain parameters and their contributions and roles they play.
· To derive effusion rate from spaceborne data using equation [3], the uncertainties in emissivity (used and an input) are directly related to lava surface temperatures derivation, which are used in equation [2]. A paragraph on heat equation has been added to clarify the control on the rate of transfer, as temperatures are raised to the forth power, thus relevant to emissivity variation.
· The bottom term in equation [3] represents the heat in the two components of the flow, liquid and solid, using an estimate for bulk density of the flow. ρ, CP, CL, Φ are specific values for this particular lava type and are typical for Mt Etna. The most influential parameter that could significantly change the result, if varied, is , with value range between 100 to 200 K. For the purpose of our study, focussing on emissivity variation, all the parameters at the bottom of equation [3] were treated as constants, including (set to 150 K) and Φ (set to 0.45) for each of three emissivity calculations (0.8, 0.93 and 1.0). Therefore, the change in emissivity (and its uncertainty) will play the primary role (indirectly through equation [2]) in respect to other parameters in equation [3].
Reviewer #1: 243-244: I calculated a 20% emissivity change affects only for a 2% and 6% the temperature estimates of SWIR and TIR bands respectively. It seems from your results that emissivity does not play an important role in estimating temperature.
Authors:
· Figure 6 is an example in which 20% emissivity change (i.e. variation between 0.80 and 1.0) applied uniformly to RS data produced 15 K and 25 K temperature difference in SWIR and TIR bands respectively. 2-6% difference in calculated integrated pixel temperatures on the ground from RS is a simplified approach, which assumes that emissivity does not change as a function of wavelength or the temperature. It is however evident from Figures 2 & 3 that emissivity is wavelength dependent.
· To enhance the ‘static’ ASTER GED and low-temperature emissivity argument, we added a high-temperature FTIR results (400-900 K) to the Method and Results Sections. Our preliminary results from high-temperature FTIR analysis suggest that emissivity changes significantly with temperature. This emissivity/temperature trend can be used as a guide in future modelling and spaceborne applications to derive a reliable and exploitable predictive emissivity at a range of temperatures and wavelengths for Etnean ‘aa’ lavas. This approach may improve emissivity assessment and give more profound insight into role emissivity to estimate lava surface temperatures more accurately, using high-spatial resolution RS data. Similarly, using our emissivity/temperature trend in tandem with numerical modelling (including all the thermal components), might better account for the changing thermal conditions and rheology of the flow.
Reviewer #1: 270-277: I calculated an Lmax increase of only 5% (300m/6400m) and not of 11% as stated. This paragraph is not clear. Please re-write it.
Authors:
· Lmax increase has been changed to 5%. Paragraph has been rewritten for clarity.
Reviewer #1: 313-317: Why the best results in terms of area covered (1.94 km2) and average distribution (12.4 m) are obtained by the MAGFLOW simulation run with an emissivity of 0.8, while the best fit for length (6.4 km) is obtained by the simulation run with ε = 0.93? Please, discuss this issue.
Authors:
· Having almost the same extent does not guarantee a best fit (areas may not overlap wholly or in part). We redrafted the paragraph to better explain the results.
Reviewer #1: 338-339: It seems that a 20% emissivity change would affect distance-to-run estimation for only 5% (300m/6400m). This is a good result, which would rather suggest that emissivity uncertainties does not impact very much on hazard mitigation and planning.
Authors:
· Our ‘simple’ (uniform emissivity) results in Figure 6 may suggest that. However, by applying our emissivity/temperature trend method (described above) may improve lava surface temperatures estimate (more emissivities within the same pixel, due to a range of temperatures in the same pixel) and give more profound insight into the role of emissivity to ‘distance-to-run’. If validated, this kind of spaceborne support to volcano monitoring could prove to be important for hazard mitigation and planning. A difference of only few hundred meters can make a substantial impact on hazard mitigation action, especially in areas of high population density.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Author of manuscript: Nikola Rogic et. al.
Manuscript ID: remotesensing-447306
Title of manuscript : Role of emissivity in lava flow ‘distance-to-run’ estimates from satellite-based volcano monitoring
Comments and/or suggestions to the author:
The authors investigated the relation between lava emissivity and lava flow ‘distance-to-run’ and found that decrease 0.2 emissivity results increase 500 m ‘distance-to-run’ at Mt. Etna. However, this result is reasonable and not surprising.
Emissivity considerations both laboratory and satellite data (lines from 79 to 200) are inadequate for this paper. Generally, melted lave temperature is more than 1000 K, the authors should consider lava emissivity at melting temperature. Samples should be collected uniformly in the whole lava flow area. Why the samples are clustered into two groups?
The authors should revise their paper.
Author Response
Manuscript Ref. remotesensing-447306: “Role of emissivity in lava flow ‘distance-to-run’ estimates from satellite-based volcano monitoring” by Nikola Rogic, Annalisa Cappello, Fabrizio Ferrucci
Dear Reviewer # 2,
We accepted and inserted in the revised version all the corrections and you suggested. We very much appreciate your constructive criticism, which helped us improve the manuscript.
A point-by-point response to the concerns you raised is provided below. An annotated copy of the manuscript, where all the changes are track-coloured, has been uploaded.
We report the changes made on the manuscript following the referee’s suggestions:
Reviewer #2
The reviewer commented: “The authors investigated the relation between lava emissivity and lava flow ‘distance-to-run’ and found that decrease 0.2 emissivity results increase 500 m ‘distance-to-run’ at Mt. Etna. However, this result is reasonable and not surprising.”
Reviewer #2: Emissivity considerations both laboratory and satellite data (lines from 79 to 200) are inadequate for this paper. Generally, melted lave temperature is more than 1000 K, the authors should consider lava emissivity at melting temperature. Samples should be collected uniformly in the whole lava flow area. Why the samples are clustered into two groups?
Authors:
· The additional high-temperature FTIR results (400-900 K) have been added to the Method and Results Sections to reflect emissivity trends with temperature increase and to quantify the impact on lava flow ‘distance-to-run’ estimates
· Samples have been collected uniformly right across the flow. Grouping was mainly to assess samples local heterogeneity. Considering sample’s compositional homogeneity, results for all samples are now presented as a Series and referred to as NRE.4S. The range (variation) for samples in NRE.4S is indicated either in the caption or by error bars (or both) in results presented for laboratory and spaceborne data.
Author Response File: Author Response.pdf
Reviewer 3 Report
This was an interesting paper, with a a lot of promise. The only real issue I have is the relation of the causal effects. Effusion rate does not change emissivity, or vice versa. Rather changes in apparent emissivity represent changes in the surface condition of the lava, and likely the liquid to solid transition. This is useful information to determine the dynamics of the flow, and when it builds a crust and when the radiative heat loss drops to a level that matches the convective heat loss. I would rethink the causal relationships, add a few references that I've given, and then it would be just fine.
Comments for author File: Comments.pdf
Author Response
Manuscript Ref. remotesensing-447306: “Role of emissivity in lava flow ‘distance-to-run’ estimates from satellite-based volcano monitoring” by Nikola Rogic, Annalisa Cappello, Fabrizio Ferrucci
Dear Reviewer # 3,
We accepted and inserted in the revised version all the corrections and you suggested. We very much appreciate your constructive criticism, which helped us improve the manuscript.
A point-by-point response to the concerns you raised is provided below. An annotated copy of the manuscript, where all the changes are track-coloured, has been uploaded.
We report the changes made on the manuscript following the referee’s suggestions:
Reviewer #3 provided a PDF document with a detailed feedback
Reviewer #3: PDF document comments
Authors:
· We removed Table 1 and listed types of analyses used in this study in the text.
· The FTIR set-up is detailed in Maturilli et al., 2018, so diagram of the set up was not included due to the space limitations.
· We included additional references as suggested.
· Emissivity does change with temperature, so additional high-temperature FTIR results (400-900 K) have been added to the Method and Results Sections to enhance argument on emissivity/temperature variation uncertainty.
· Both equations [2] and [3] have been adjusted and explained accordingly.
· The suggested interpretation for using change in emissivity as a proxy for the degree of crusting of the lava flow has been discussed, which we found to be a very helpful addition.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
The authors answered all my comments and modified the text in agreement. I have no other comments to the text
Reviewer 2 Report
Author of manuscript: Nikola Rogic et. al.
Manuscript ID: remotesensing-447306
Title of manuscript : Role of emissivity in lava flow ‘distance-to-run’ estimates from satellite-based volcano monitoring
Comments and/or suggestions to the author:
The authors investigated the relation between lava emissivity and lava flow ‘distance-to-run’ and found that decrease 0.2 emissivity results increase 600 m ‘distance-to-run’ at Mt. Etna. This manuscript is interesting for readers of Remote Sensing.
