Sea Surface Salinity Subfootprint Variability from a Global High-Resolution Model

Round 1

Reviewer 1 Report

 

Review of “Sea surface salinity subfootprint variability from a global high-resolution model” by Bingham et al.

 

This paper uses output from a very high resolution (1/48 deg) simulation to examine subfootprint variability (SFV) of sea surface salinity in the context of satellite salinity measurements. Seasonal and spatial patterns of SFV are examined, focusing on 40 km and 100 km footprint sizes (corresponding to the Aquarius and SMAP satellites). The paper is fairly descriptive but also provides a nice discussion of the observed results. The paper is clear and easy to read, and is an important addition to the field of satellite remote sensing of salinity: specifically, understanding the uncertainties in satellite measurements. I recommend publication after the following points are addressed.

 

 

Main points:

 

1. Since the effective horizontal resolution of the model is 10 km (L109), could the authors comment on how realistic the estimates of SFV over 40 km scales are? What about 20 km scales (Fig 4,5)?

 

2. In the discussion (L347-357), the authors state that seasonal SFV is caused by either rain or submesoscale variability. However, large-scale fronts (such as found near western boundary currents) and mesoscale features also generate significant SFV, so seasonal patterns in SFV could presumably result from seasonal variability on scales larger than submesoscale (e.g., excursion of WBCs, variability of mesoscale features). It may not be surprising that in regions with strong large-scale fronts (such as the Kuroshio extension, L352) the seasonal patterns of SFV are not consistent with the seasonal patterns of submesoscale turbulence observed by other authors. I suggest expanding your discussion to mention other sources of seasonal SFV.

 

 

 

Minor points:

 

• L16: replace “numerical model, the MITgcm” with something like “numerical model simulation of the MITgcm”, since the MITgcm has many flavors and not all are 1/48deg

• L88-89: perhaps extend the statement that models are useful for estimating SFV in regions without heavy rainfall with a comment that models are most useful in regions where ocean dynamics generate the SSS variability

• Section 2.1: mention the vertical resolution of the upper level of the model (i.e., the one you use to represent SSS)

• L123: clarify whether “20 km for SMAP” refers to the footprint size or d_o. Is the example in Fig 1 for Aquarius? If so, perhaps say that explicitly.

• L125-127: is there a reference for the 50%/44%/6% breakdown stated here?

• L130 please state the value of C for each of the footprint sizes you use

• L148: suggest “for which weights, w_i, are greater than 0.5.”

• Fig 2: suggest adding at least one value on the colorbar between 0.005 and 0.1, as it’s difficult to visually interpolate the log scale

• Figure 2, 4: it’s unclear why you refer to the regions of interest as “random boxes”, since they seem carefully selected to highlight interesting locations. Perhaps “Regions of interest”? You might also label regions with their names for easier interpretation of Fig 4 (so the reader doesn’t have to keep flipping back to Fig 2a)

• Fig 4d please extend y-axis to include the full range of the Box G line

• Fig 6 – I like the concept but the caption is a bit confusing so I don’t understand what the figure is showing. What does “evaluation points” mean? “The normalized area … of SFV is maximum in a given month” does not seem grammatically correct – do you mean something like “the normalized area that SFV has its maximum value”? If so, what maximum value do you use? Does this analysis basically pick out the area of strongest SFV (e.g., western boundary currents in the higher-latitude bands)?

• Fig 2 c/d, Fig 9: suggest normalizing the histograms by total # of boxes so these results could more easily be compared to other analyses. Or, at least state the total # of boxes.

• L326: replace “version of the MITgcm” with “MITgcm simulation”

 

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

The purpose of this study is to determine the sub footprint variability of sea surface salinity by using a global high-resolution model for different season. It estimates the sub footprint variability to include in error budgets for satellite. Sea surface salinity measurement is important for the development of the global hydrological cycle.  There have been a number of studies of the seasonality of sub mesoscale variability in the ocean as it relates to such quantities as eddy kinetic energy. This is highly useful for the aquatic or oceanographic studies

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Summary: The manuscript titled "Sea surface salinity subfootprint variability from a global high-resolution model" by Bingham et al. sets out to describe and characterize the subfootprint variability (SFV) of sea surface salinity (SSS) using a high-resolution model.  While describing the SFV of SSS has been done in the past, mainly at local scales, this furthers our understanding of SFV by further characterizing the seasonality of SFV and its dependence on footprint size at a near-global scale.  The manuscript is well written and its results on SFV of SSS will certainly be welcomed by the salinity and hydrological cycle community; however, I feel some revisions should occur before publication.

 

Major Comments:

General: The discussion on the seasonality of SFV is lengthy and at times repetitious, while the discussion on RE and its implications is incomplete.

1a.)  A majority of the manuscript focuses on the seasonality of SFV for SSS, but there is a lack of evidence of seasonality in the main text aside from discussions/focus on min/max time periods (fall/spring).  Of course, if we were to examine the SFV for each month in the supplement this would be evident, but in my view there should be evidence of seasonality in the main text.  For example, a simple figure illustrating an area average time series of SFV for a few different areas would suffice.

1b.)  This is related to 1a (above), but did the authors consider illustrating the seasonal cycle for both 100 and 40km footprints by applying a harmonic analysis to SFV?  In this way, a measure of the amplitude and phase of SFV could be shown and would help condense the lengthy discussion on the seasonality of SFV.  Do most regions undergo an annual cycle, or do some regions experience a semi-annual cycle?  This is not clear in the present text.

2.)  Based on the Introduction, one of the motivators behind this manuscript was to better quantify and understand representation error (RE) using a high resolution model.  Many validation experiments have been carried out between satellite sea surface salinity and in situ salinity measurements (some are referenced in this manuscript).  However, the results of RE and a discussion relating it back to these validation efforts and its implications is lacking in my view.  For example, we know Aquarius had a mission goal of measuring salinity to an accuracy of 0.2; how do these results of RE relate to that and the validation efforts that went into comparing Aquarius and in situ?

 

Minor Comments:

Figures 2, 3 and 8: What is the reason for the logarithmic color scales?  Is it necessary to show areas in different shadings that experience SFV less than 0.1 (Figures 2/3)?  Also, I would recommend moving away from the jet color map and using something more color blind friendly.

Lines 399/340: There appears to be some words missing with the sentence that begins with "In some..."

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 3 Report

The authors adequately addressed all of my concerns in their revision.  I only have a single minor revision recommendation and that is related to the time span of the model.  As the authors described in their comments, a harmonic analysis of a single year's worth of data would be quite noisy.  Although the authors cite various publications that diagnose SSS SFV in various regions, and many of them support the model findings, I would suggest that the authors still include a bit more of a "disclaimer" in that this is still only one year's worth of high-resolution model data and the seasonal cycle that was diagnosed could be subject to changes when more years of data are incorporated.  I know this is touched upon, but reminding the reader about this is important in my view.  Overall, this manuscript was well-written and will be well-received!

Author Response

Please see the attachment.

Author Response File: Author Response.pdf