RADAR-Vegetation Structural Perpendicular Index (R-VSPI) for the Quantification of Wildfire Impact and Post-Fire Vegetation Recovery

Round 1

Reviewer 1 Report

 

A well written and significant contribution to research on satellite detection of vegetation change as a result of fire. The paper does not provide a rigorous ground based assessment of structural changes in vegetation resultant of fire. Rather it offers control from similar patches of unburnt vegetation in proximity to the fire. This is very much understood, as both cost and practicality of ground based truth assessments make this approach very difficult to implement. A lack of a deeper discussion of the impacts of rain on the radar based metrics, is one area in which the paper could be improved. Other minor issues are:

Lines 135 – 136: Stating that Sentinel 1 provides 10m resolution and Sentinel 2 provides 20m resolution respectively, seems to ignore that S2 records 4 spectral bands at 10m and S1 spatial resolution is reduced when speckle is removed. At best they are similar at 20m resolution

Line 332: ( is missing a corresponding  parenthesis

Line 365: should RFDI be mRFDI?

Figure 1A, Figure 4B and Figure 8 – X and Y Axes show DN values – these should be reflectance values. The DN are probably scaled reflectance values, but this should be explained in the captions to figures.

Figure 7: explain the curved thin black line which is superimposed on all images – is it a road? Also discuss the vertical band highly evident in the VV/VH ratio images, but also evident in other radar based images

Lines 412 – 413: NDVI indicating recovery is strongly associated with regrowth – this can be in both epicormic growth but also ground covers including invasive weeds, which are not representative of pre-fire conditions. There is some reference to this at lines 457 / 458.

Line 464: “Using R-VSPI and VSPI in combination …. may be advantageous”. This concept should be further explored – it is discussed again in line 484 – but it would be an improvement if the paper could suggest a mechanism to combine the indices.

 

Line 508: make reference to NiSAR (S and L band) and Biomass (P Band) in particular.

 

 

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

In this paper the authors have proposed a SAR based index using polarization channels from Sentinel-1 for quantifying changes in fuel load due to wildfires. The presented results shows that the proposed indexes (R-VSPI and VSPI) area able to estimate the wildfire-induced forest changes as the orthogonal distance from a linear reference line that characterized the undisturbed forest. In my opinion, this paper addresses an interesting topic and offers some innovation. The manuscript is definitely worth publishing.

Herewith I have attached my minor comment to address.

Comments for author File: Comments.pdf

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 3 Report

 

Review of the paper: RADAR-Vegetation Structural Perpendicular Index (R-VSPI) for the quantification of wildfire impact and post-fire vegetation recovery

 Authors: Aakash Chhabra, Christoph Rüdiger, Marta Yebra, Thomas Jagdhuber, James Hilton

 Paper id:  1750075

 General comments

The paper analyses the potential of remotely sensed Sentinel-1 backscatter data as an alternative to optical-based observations to achieve an accurate and timely measurements of wildfire fuel characteristics. The study is carried out on the bases of a simple approach, which exploits the polarimetric information available from the SAR data. The use of polarization features of radar response at C band to estimate vegetation and soil characteristics is not a novelty and the papers basically use tools available from previous studies, referencing to a satisfactory bibliography. The main value of the paper is the detailed and well described data analysis, based on a heterogeneous data set and corresponding ground truth which has not often present in similar studies. In general, the paper is well written and clear. I have two comments/suggestions.

In some parts of the manuscript (e.g. 178-181) the authors underline a consolidated statement “The backscatter from VV is typically sensitive towards surface scattering components (such as the landscape moisture), whereas the energy measured from the VH…..is related with volume scattering in forests [53]”. Actually, this statement is very basic: the interaction with the different polarization components is a complex issue and a huge of studies are present in literature about the influence of geometrical and dielectric characteristics of the vegetation layer. This fact, of course, does not invalidate the positive results experimentally obtained with the proposed indexes but probably demands a more exhaustive study with different typology of vegetation to extend the actual operativity of the methodology worldwide; just to give an example, I guess that the application to Mediterranean scrub fires could be not straightforward. The second comment concerns the suggestion to evaluate the use of other polarization tools, as for example the calculation of the different dual polarimetric descriptors based on entropy/alpha decomposition (Cloude & Pottier An entropy based classification scheme for land applications of polarimetric SAR. IEEE TGRS 35(1):68–78 1996) to improve the data analysis.

In 220 reference [40] does not contain details of the speckle filter; the following for example is more exhaustive: J. S. Lee, I. Jurkevich, P. Dewaele, P. Wambacq and A. Costerlinck, "Speckle filtering of synthetic aperture radar images: A review", Remote Sens. Rev., vol. 8, pp. 311-340, Apr. 1994.

Author Response

Please see the attachment.

Author Response File: Author Response.docx