Field Testing Satellite-Derived Vegetation Health Indices for a Koala Habitat Managers Toolkit
Round 1
Reviewer 1 Report
This revised version is substantially improved, and worthy of publication.
I would point out only 4 (very minor) edits:
Line 223: "stakeholders"
Line 461: a stray letter "f"
line 587: "...written up in the..."
Line 769: change " ...following 12 maps..." to "...following six figures...
Author Response
The authors thank the reviewer for their suggestions - please find attached our response to the corrections requested.
Michael Hewson CSC PhD
Author Response File: 
Author Response.pdf
Reviewer 2 Report
General Comments:
The manuscript presents a study on the development of a koala habitat management ‘toolbox’ that utilizes two satellite image products i.e. Rapid Eye and Sentinel-2, to derive vegetation state proxies in order to assess vegetation health.
Specific comments:
Equation 4 contradicts Lines 201 and 202 by using TIR rather than SWIR.
Line 246 - 247: Please consider checking the spelling of ‘resampling’. In addition, reference to resampling is made with respect to Band 11 (SWIR1) only. Was resampling carried out for the RedEdge band used for NDRE computation? Which of the RedEdge bands was used? In order to explicitly indicate which bands were used, the equations for the indices i.e. equations (1), (2), and (3), should be written out in the format of equation (4), for both satellite image products.
Lines 348 – 350: Kindly provide further details on the “routines of the ESA SNAP image processing software” and put the description in the context of Lines 138 – 143. Was the atmospheric correction carried out to achieve a Level-2A image? If so, what specific process/ routine was carried out? Were additional tools/ packages required?
Line 406: Please check the spelling.
Line 408: Reference is again made to a TIR band. Please see the MSI band designations explained here: https://www.satimagingcorp.com/satellite-sensors/other-satellite-sensors/sentinel-2a/
Author Response
The authors thank the reviewer for their suggestions - please find attached our response to the corrections requested.
Michael Hewson CSC PhD
Author Response File: Author Response.pdf
Reviewer 3 Report
Comments to the Author
This paper presents a methodological approach to the Field testing satellite-derived vegetation health indices for a koala habitat managers toolkit.
I have the following comments:
- I suggest adding results obtained (at least the main ones), and main conclusions (at least the main ones) in the abstract. The method of the study is not clear.
- I suggest specifying TIR1600 in equation 4.
- I suggest describing why chose this date for four fieldwork sites in section 2.4. You can use this phrase to respond to a comment about chose of the date on page 9 "Optical instrument image acquisition dates can be variable when accounting for regional cloudiness".
- I suggest verifying and indicating assumptions for the Pearson Correlation, such as both variables should be normally distributed; verify linearity and homoscedasticity.
- In table 1, change "v" to "vs". In table 1 these values indicate that the R square is not adequate, is necessary to verify assumptions to use Pearson correlation. I suggest obtaining a p-value from correlation. A low R-squared value indicates that your independent variable is not explaining much in the variation of your dependent variable
- In fig. B, page 22, it is suggested add coordinates.
- R square values are very low (see appendix C) " A low R-squared value indicates that your independent variable is not explaining much in the variation of your dependent variable".
Comments for author File: Comments.pdf
Author Response
The authors thank the reviewer for their suggestions - please find attached our response to the corrections requested.
Michael Hewson CSC PhD
Author Response File: Author Response.pdf
Round 2
Reviewer 3 Report
No suggestions
This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.
Round 1
Reviewer 1 Report
A good application paper
Author Response
Please see the attachment.
Reviewer 2 Report
Please consider the comments in the attached PDF
Comments for author File: Comments.pdf
Author Response
Please see the attachment.
Author Response File: 
Author Response.docx
Reviewer 3 Report
General Comments:
The manuscript presents a study on the development of a koala habitat management ‘toolbox’ that utilizes satellite imagery to derive vegetation indices to assess vegetation health. In this study imagery from the Sentinel-2 mission on two dates were used to derive NDVI, NDRE, EWT, and LAI. Field measurements were also carried out to measure LAI. A major concern is that the description of the background of the study area, the sampling design, the theoretical background behind comparing VIs to LAI, and the methods used for validation, are inadequate. The sampling design and the fact that the study only took into consideration two temporal instances could explain the poor correlation results, especially between the field measured and satellite-derived NDVI. Given that the aim of the study was to develop a toolbox for both geoscience specialists and novice users, with a view of aiding rapid vegetation analysis on a continuous basis, a workflow incorporating tools for rapid acquisition, ingestion, processing, and analysis of satellite imagery via cloud computing services such as Google Earth Engine (GEE) would have been expected. Such a toolbox or application could also support time-series analysis, which could provide a better basis for assessment and monitoring with respect to ENSO events. An application toolbox with a simple graphical user interface that does not require the user to interact with the back-end processes in SNAP could prove to be far more efficient and easier to deploy.
Specific comments:
Lines 125 -129: The future tense is used. This is confusing as the reader expects that the activities described were already undertaken.
The introduction needs to include background information on the typical characteristics of Koala habitats. Questions like: what is the dominant vegetation type, what are the spatial scales, are the habitats fragmented? Should be answered. This information allows the reader to understand the choice of indices and satellite imagery used in this study and supports the assertions made in Lines 182 - 185, Lines 214 - 215, and Lines 226 - 233.
Line 198: Please consider revising the use of Thermal Infrared (TIR) in the manuscript and instead use the widely known and accepted band designations. As noted in [17], cited in this manuscript, the thermal infrared spectrum is considered to be between 6.0 and 15.0 mm. In most remote sensing literature, and specifically for terrestrial remote sensing, the spectral range of 3 to 35 μm is referred to as thermal infrared. Therefore, 1600nm, referred to in the manuscript as TIR, falls within the short wave infra-red (SWIR) range and is used as such in [17].
Lines 219 - 220: Again the use of TIR is confusing since Sentinel-2 cannot be considered to have a TIR band. Please see the MSI band designations explained here: https://www.satimagingcorp.com/satellite-sensors/other-satellite-sensors/sentinel-2a/
Line 241: Please note that the 10-meter resolution is not applicable to all bands, and especially the SWIR bands, which are used in EWT estimation. Which of the two SWIR bands was used? Was there any resolution harmonization or resampling carried out?
Lines 280 - 284: Please consider including a general study area map placing the specific area of study in the global and national contexts, preferably with a base image. In addition, the geographical coordinates of the fieldwork sites and a description of the climate and topography of the study area would add value.
Line 297: Kindly consider overlaying the locations of these sites in the maps recommended in the previous comment for context. For each of the field sites, maps showing the sampling locations would also be appreciated.
Line 370: Kindly provide details of SMAP RSM. It has been introduced here with no prior information.
Line 379/ Figure B-1: Please provide a legend. What does the purple outline represent?
Lines 394 - 397/ Table 1/ Appendix C: Please shed light on the scientific value of comparing satellite-derived NDVI and NDRE with measured LAI. In Lines 354 - 357, a procedure for satellite-derived LAI is described, along with a comparison to the LAI equation, and that parity was found. It is interesting that there is a relatively weak/ moderate relationship between the measured and satellite-derived LAI, especially for RapidEye, given the statement in Line 357.
Lines 421- 424: Given the weak to moderate relationship between the measured and satellite-derived LAI as shown in Table 1, it would be remiss to further compare the measured LAI to satellite-derived NDVI and NDRE. Further, in looking at the cited publication [22], it was not concluded that there is a strong relationship between spectral band ratios and LAI.
Figure B-3: Please consider using a more intuitive color palette or ramp. Since not the entire area is covered by vegetation, you could adopt the standard color ramp for ArcGIS (NDVI3), where "Red and orange pixels represent areas with no vegetation. Yellow pixels represent areas with low to moderate vegetation. Green pixels represent areas with high vegetation density and vigor." In addition, the legend needs to be revised. It appears to be the direct output of the software that relays no meaningful information to the reader e.g. what do 'value' and 'dimensionless' mean?
Author Response
Please see the attachment.
Author Response File: Author Response.docx
