Sentinel-1 Backscatter Time Series for Characterization of Evapotranspiration Dynamics over Temperate Coniferous Forests

Round 1

Reviewer 1 Report

The manuscript compares Sentinel-1 backscatter time series and evapotranspiration time series over a forest region in Germany. A high correlation between the two variables was observed, especially after nonlinear filtering to extract seasonal components. The only weak part of this paper is that the underlying mechanism of how transpiration affects backscatter is thus far not well understood, risking a correlation vs causation fallacy. Nevertheless, this is an actively explored topic with potentially significant implications, undoubtably making this study a worthwhile contribution.

The manuscript is well written, and care has been taken in handling the data, making it a very interesting read. Under the circumstances of our current knowledge of the topic and the available data, this is a very good manuscript and I recommend it for publishing after minor revisions.

 

Although it might not be feasible for the current manuscript, I would encourage the authors to strengthen their argument regarding the connection between backscatter and (evapo)transpiration. Transpiration is a flux which, under certain conditions, has been observed to cause a decrease in the water content of plant structures under observation (and usually within reach of human arms). Why then do the results in this paper show an increase in backscatter during high transpiration periods? The underlying mechanism is difficult to relate to radar observations, but this does not mean one should not try, especially since multiple factors can contribute to a seasonal variation in backscatter. It is evident that the authors have not considered literature from the tree physiology community since none of it is cited. Doing so would significantly strengthen the authors’ arguments.

 

Specific comments:

 

P1, L 106:

I think “input variables” is more suitable than “parameters” in this context, highlighting the method’s simplicity.

 

Figure 1:

Please label the axes. I assume they should be “Northing [m]” and “Easting [m]”.

Alternatively, remove tick numbers since a scale already given.

 

P5, L 184:

Change “wavelength” to “frequency”.

 

P5, L 184:

Change “ground range” to “ground range detected”.

 

Figure 2:

The line colors in the legend are indistinguishable when printed, although this is a minor issue.

 

Figure 2 caption:

Instead of specifying the averaging window size in number of samples, I recommend specifying it as a time duration. A window size of 10 is meaningless on its own. Please spare the reader the discomfort of having to look up the sampling interval.

 

Equations 2-4:

Avoid using ∗ as a symbol for multiplication. Use ×, ⋅ or no symbol.

 

P 11, L 377-380:

I recommend removing this section since the differences in variabilities upon which this argument is based are insignificant.

 

P 11, L 419-420:

This statement is too strong. The cited article deals with a radiometer operating at L-band and matches VOD variations to water content/potential variations via a simple model with little regard to tree physiology. I recommend rather stating something like “… is not directly sensitive to temperature but rather water content variations driven by environmental changes in, among other variables, temperature.”

 

P 12, L 457:

Remove “a”.

 

P 13, L 483-487:

Again, if water is lost by these small scatterers to which VH is sensitive to, why is a higher backscatter observed during periods of high transpiration? Currently it is not clear where in the tree water is lost and gained at different times of the day, especially higher up in the canopy where very few studies have been conducted thus far. Water is indeed lost to the atmosphere in leaves/needles, but it is not yet clear what impact this has on the water content within these structures. I recommend reformulating this sentence so that the argument is clear and honest. Reference [58] might have helpful information.

 

Table 3 caption:

Were ET time series also filtered? If so, make this clear in the caption.

 

Figure 7 caption line 3:

Add “ Sentinel-1 time series” after “SSA- and SMA-filtered”.

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Please see the attached file.

Comments for author File: Comments.pdf

Author Response

Please see the attachment.

Author Response File: Author Response.pdf