the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 19 Jul 2022
Climatic control of the surface mass balance of the Patagonian Icefields
Tomás Carrasco-Escaff
Maisa Rojas
René Garreaud
Deniz Bozkurt
Marius Schaefer
Abstract. The Patagonian Icefields (Northern and Southern Patagonian Icefields) are the largest ice masses in the Andes Cordillera. Despite its importance, little is known about the main mechanisms that underpin the interaction between these ice masses and climate. Furthermore, the nature of large-scale climatic control over the surface mass variations of the Patagonian Icefields still remains unclear. The main aim of this study is to understand the present-day climatic control of the surface mass balance (SMB) of the Patagonian Icefields at interannual timescales, especially considering large-scale processes.
We modeled the present-day (1980–2015) glacioclimatic surface conditions for the southern Andes Cordillera by statistically downscaling the output from a regional climate model (RegCMv4) from a 10 km spatial resolution to a 450 m resolution grid, and then using the downscaled fields as input for a simplified SMB model. Series of spatially averaged modeled fields over the Patagonian Icefields were used to derive regression and correlation maps against fields from the ERA-Interim reanalysis.
Years of relatively high SMB are associated with the establishment of an anomalous low-pressure center near the Drake Passage, the Drake low, that induces an anomalous cyclonic circulation accompanied with enhanced westerlies impinging the Patagonian Icefields, which in turn leads to increases in the precipitation and the accumulation over the icefields. Also, the Drake low is thermodynamically maintained by a core of cold air that tends to reduce the ablation. Years of relatively low SMB are associated with the opposite conditions.
We found low dependence of the SMB on main atmospheric modes of variability (El Niño-Southern Oscillation, Southern Annular Mode), revealing a poor ability of the associated indices to reproduce interannual variability of the SMB. Instead, this study highlights the Drake Passage as a key region that has the potential to influence the SMB variability of the Patagonian Icefields.
-
Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
-
Preprint
(12214 KB)
-
Supplement
(4662 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(12214 KB) -
Supplement
(4662 KB) - BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Tomás Carrasco-Escaff et al.
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2022-603', Anonymous Referee #1, 30 Aug 2022
Review of “Climatic control of the surface mass balance of the Patagonian Icefields” by Carrasco-Escaff et al., submitted to The Cryosphere.
The authors present an interesting study that adds valuable new knowledge to climate and glacier science related to southern South America. The study has been carried out well and is sufficiently documented over large parts of the manuscript. Just the description of the sensitivity analysis is in parts hard to follow and some efforts should be undertaken to improve readability of this section. Apart from this, I have two major objections that prevent me from supporting publication of the article in its present form:
Major comment 1)
The downscaling of solar radiation as it is described in the one sentence provided in L183f has to be questioned. Bilinear interpolation of shortwave radiation on a non-systematically varying surface (like a DEM representing natural terrain) leads to wrong values at the higher-resolution scale. The angle between incoming direct solar radiation and surface slope/aspect (incidence angle) is crucial in determining the right amount of energy reaching the glacier surface. Hence, simply interpolating radiation values from low- to high-resolution grids introduces errors that could easily double or halve solar radiation energy reaching the surface. Regarding diffuse radiation, the skyview factors of the high-resolution grid cells might probably differ considerably from those of low-resolution fields. Taken together, it requires more to downscale solar radiation than just bilinear interpolation.
As spatiotemporal variability of solar geometry can easily be implemented in a downscaling model, the approach needs to be refined by considering incidence angles at each grid cell of the high-resolution topography. Otherwise, the resulting values are simply wrong. Moreover, a validation needs to presented that compares original and downscaled values to in situ measurements (ideally at an on-glacier weather stations). Such a validation must also be presented for T and P, as otherwise it is hard to argue why the RegCM fields can be used for reliable SMB modeling, especially as they show considerable biases to the reference CR2MET climate, which are corrected in a rather simple way only. I’m sure that the team of authors has access to such data even if it might cover only a short period of time.
These validations might also help to overcome the problem of validating the modeled SMB with respect to inter- and intra-annual variability. Assuming that downscaled T, P and R clearly show seasonal variability on a local scale, this would also suggest that the modeled SMB might be reliable in this respect.
Major comment 2)
Climate forcing is analyzed using the SMB integrated over NPI and SPI together. This spatially undifferentiated way of looking at the outcome of this study is a missed opportunity that should be accounted for in a revised and extended version of the study. In its present form the analysis prohibits to get an idea about potential regional variability of forcing mechanisms across Patagonia. I would like to see similar figures to Figs. 6-11 be added to the supplement that show the correlations with only NPI and SPI. Analyzing the differences of these two sets of maps/graphs would give valuable insight into regional variations of climate forcing across Patagonia. This would strengthen the interpretation of the so far presented results which just integrate over NPI and SPI. Sections 3.3-3.5, as well as discussion and conclusion should then be extended accordingly. As we know from the literate that NPI and SPI do not always show the same patterns of glacier change, such an analysis might be of really high value to science – even if it shows that climate forcing mechanisms do not differ significantly for NPI and SPI.
In addition to these comments I have quite some minor comments that also needs some attention of the authors. Based on the two major comments above and the minor comments below, I suggest to return the manuscript for major revision.
Minor comments:
L9: better: ...fields of climate variables from the ERA-Interim…
L40: These positive trends fit to the recent southward shift and strengthening of the southern hemispheric westerly wind belt (e.g. Goyal et al. 2021, doi:10.1029/2020GL090849), which might be of interest here.
L55-57: These moister than average conditions in southern Patagonia have already been suggested to significantly influence SMB (Möller et al. 2007, doi:10.3189/172756407782871530), which should be noted here.
L80: better: …, i.e. the net change of mass at the surface, … “Gain” suggest an increasing mass of ice, but SMB has been positive and negative in the period studied. See Cogley et al. 2011 (Glossary of Glacier Mass Balance) for further details on the related terminology.
L81ff: I see no need to explain glacier mass balance in such detail as the manuscript is written for the cryosphere-centered journal. E.g. basal melting should only be mentioned if it is of interest at the glaciers modeled in the presented study.
L95ff: Braun et al. 2019 and Dussaillant et al. 2019 (both in the manuscript) should also be mentioned here. And it should be discussed that these two remote sensing studies have shown strong mass loss especially over the SPI, which contrasts the positive SMB mentioned before. In its present form the reader gets a picture of increasing ice masses in southern Patagonia, which is wrong.
L129: Why ERA-Interim and not ERA5 which is available for quite a while now?
L134: Also provide reference to Alvarez-Garreton et al. 2018 here, and not only at the end of the paragraph.
L132-140: What makes the CR2MET dataset a reliable reference? I do not question here that it could be used as this, but I would greatly appreciate additional argumentation. It is necessary to outline and explain how well this dataset represents in situ conditions. Moreover, information about shortcomings and especially inaccuracies of the dataset are needed to be able to judge about its reliability. And finally (maybe most important) why are the RegCM fields created and used when CR2MET already exists? What is the advantage of RegCM over CR2MET and does this advantage justify the introduction of additional uncertainty (by comparing it to CR2MET before usage)?
L147: better: “… of world-wide glacier extent at the beginning…”, as “extension” implies a process of increase rather than a static condition
L158: not clear what is meant here: “Lastly, we spatially unweighted averaged the meteorological forcing…”
L159: better: “Only grid points within…” (omit “Note that”)
L192: provide reference for this representation of the fraction of solid precipitation
L209ff: It would be interesting to get some values on the distribution of snow/firn after the spin-up time: Give average numbers for snow-/firnline altitudes across the study area and discuss potential spatial variations in case they exist. Give reference to other studies which derived snowline altitudes in Patagonia and shortly compare your results to these findings.
L231-235: This is a really nice idea. However, I strongly request that also information about the bias in SMB compared to the reference SMB is somehow incorporated in the Taylor diagram (e.g. by scaled sizes or color-scales of the points shown). The so far given information about correlation and standard deviation only give insight into how well the variability is represented, but do not tell anything about resulting biases.
L239-249: This is an interesting approach, but more information is needed here. First, give reference to studies that introduced or at least support your idea. Second, give more details on how you determined the variability in the dataset and how you subsequently removed it. Also here, a quantification of biases is needed in addition to the measures of variability.
Fig. 5: I suggest to add a thin black line representing a zero SMB in the upper panel of the figure. This would increase readability and make positive and negative SMB years more easily distinguishable.
Table 2: Add information about the period represented by the given numbers to the caption.
L287: The fact that annual insolation shows a higher correlation to SMB than annual temperature further supports my initial request regarding a refined handling of solar radiation during downscaling.
L304ff: Isn’t that a necessary result of the over-simplified radiation downscaling that has been applied? I mean, how can a local-scale control over the SMB can be present when the applied downscaling is not able to produce the requited local-scale variability? (see my initial major comment) This analysis/interpretation must be redone after the radiation downscaling has been improved.
L307-318: It now entirely clear what was done here. A linear regression results in intercept and slope of a regression line, which are both important for interpretation. However, this full information is missing in Table 4 and has to be added. It must also be included in the following discussion.
L325ff: Why is solar radiation not considered here?
Figs. 6b/7b: I recommend not to use red/green colors for the isolines as these colors are hard to differentiate for a lot of color-blind people.
L410ff: It would greatly strengthen the findings of the study if comparisons to other long-term SMB time series at other Patagonian glaciers would be given. E.g. Möller & Schneider 2008 (doi:10.3189/172756408784700626) present a modeled SMB time series for Gran Campo Nevado ice cap south of the SPI. This time series e.g. shows the same strongly positive anomalies of SMB in 1990 and 1995, which supports the presented findings for SPI by showing that they fit nicely into the picture presented by other studies. Further south (e.g. Tierra del Fuego) other SMB pattern prevail (e.g. Buttstädt et al. 2009, doi:10.5194/adgeo-22-117-2009), suggesting a southward limitation of the regional pattern.
L418: Doesn’t this contradict the results that you presented before (see my comments on L287 and L304ff)? This should be clarified either here and/or above.
L418-426: This paragraph would benefit from some references to either figures or tables.
L456ff: References to other studies dealing with this or comparable issues would support your speculation and should be added and discussed shortly.
L474: This thought has not come to my mind until now: Is there any significant interannual variability in solar radiation? Or is it largely time-invariant? I’m asking because of the frequent presence of clouds in Patagonia. If there is no significant interannual variability, it would be a necessary consequence that SMB variations show almost not dependence on it. This needs to be analyzed (and outlined in the results section) before giving this broad statement, in order to potentially put it into the right context.
L490: “SBM” needs to be corrected to “SMB”
Citation: https://doi.org/10.5194/egusphere-2022-603-RC1 - AC1: 'Reply on RC1', Tomás Carrasco-Escaff, 17 Oct 2022
-
RC2: 'Comment on egusphere-2022-603', Anonymous Referee #2, 31 Aug 2022
Summary: this paper describes the climatic controls on the surface mass balance (SMB) of the North and South Patagonian Icefields (NPI, SPI). This is achieved by estimating the annual and seasonal SMB with a simple snow, firn and ice accumulation and ablation model, subsequently regressing the SMB-anomalies time series to a suite of local, regional and climate indices. Results indicate that winter precipitation and summer temperature anomalies are the main drivers of SMB interannual variability. Also, the authors find that a pressure anomaly over the Drake’s passage (the Drake low) is the dominant feature related to SMB departures, seemingly driving increased westerly winds and cooler conditions off the coast of Patagonia. No significant correlation was found between the SMB and major climate indices such as ENSO, which confirms previous work published in the area.
General comments: this is a well written paper, and is a nice contribution to the understanding of the NPI and SPI present-day behavior. The authors have taken preemptive actions to prevent the inevitable modeling uncertainties from affecting their conclusions, by focusing on correlations/anomalies only and by ensuring that potential biases in the meteorological forcings of the SMB model don’t result in major changes in the year to year variability, measured through correlation and standard deviation of the time series. The organization of the manuscript is very intuitive and the use of English language is appropriate but for a few minor issues. Because the analysis rests so strongly on the simulated mass balance, the manuscript should devote a bit more space to discussing the calibration of the four main parameters of the model, namely the threshold at which precipitation falls as snow (here set as 2°C), and the ablation parameters (albedo, c_0 and c_1). The sensitivity of the model to these parameters should in turn influence the interplay between precipitation and temperature during the accumulation season, and the relative influence of radiation and temperature during the ablation season. It may be that the main conclusions don’t change with respect to what is shown in the current version, but so far the paper seems to gloss over this topic in a manner too succinct.
Specific comments:
L132: It is not clear to me what the verification of RegCMv4 against CR2MET intends to achieve. There are clear biases shown in Fig2, which could result from several factors. Because you have threshold term in accumulation that depends on T, this bias in temperature could have compounded effects on the simulated SMB correlations. Do you do anything after verifying the two products against each other?
L201: snow, firn and ice should not be called “soil”. Please use something like “land cover”.
L210: in modeling parlance, “true” has a very specific meaning. Please revise.
L218: See general comments regarding the detail that is needed about the SMB model calibration process. Also, why do you compare the 2000-15 simulation with the 2000-19 Minowa estimates? Is it not possible to compare a common period?
L229: These biases could also result from inadequacies in the CR2MET product. In particular, if it is station-based, previous research has shown that meteorological station data in Patagonia is unreliable, particularly precipitation.
L231: a similar analysis could be performed by perturbing some of the model parameters (see general comments).
L260: If I understand correctly, AAO was calculated only for the 1979-2000 period? But SMB is available until 2015? Maybe it’d be useful to have a summary table with all datasets used, indicating time window, time-step, and citation.
L272: please reword to remind the reader that all these numerical quantities are estimates from your model. Also, the fact that annual SMB is positive means that for the ice fields to be in equilibrium (or decreasing in mass, as the literature suggests) then calving should account for the excess mass. Is that right? Also: there appears to be a slight increasing trend in the simulated SMB? Could you comment on this?
L298: How do you interpret the fact that although insolation shows the second-highest correlation with annual SMB (line 287), then the local-scale control indicates exactly the opposite? Maybe I’m missing something, but these two results seem inconsistent. Please clarify. Is this result sensitive to assumptions regarding the seasonal evolution of snow, firn and ice albedo? Nevertheless, it is expected that, unlike glaciers in mediterranean regions, solar radiation should have a minor role compared to temperature in Patagonia. High relative humidity and high very persistent cloud cover are coherent with this result.
L406: a small detail: probably “good” is not an appropriate adjective for describing correlation.
L446: suggest replacing “maintains” with “remains”.
Citation: https://doi.org/10.5194/egusphere-2022-603-RC2 - AC2: 'Reply on RC2', Tomás Carrasco-Escaff, 17 Oct 2022
-
RC3: 'Comment on egusphere-2022-603', Anonymous Referee #3, 13 Sep 2022
General comments
The study analyzes the control of the present-day climate on the surface mass balance of the Patagonian Icefields. The main goals of the study are clearly formulated, and the study is well structured and written over largest parts. In the Discussion section, a stronger comparison with and discussion of results of other SMB studies in the region could strengthen the findings. Overall, the study adds valuable knowledge to the understanding of the interaction between climate and glacier mass balance in the southern Andes. I have one major comment which needs addressing and revision before publishing the article, together with several minor comments.
Major comment
The main limitation of this study is the fact that there is no spatial analysis for the correlation of SMB with large-scale indices and climate. It is possible that e.g. SAM does have an important impact on the SMB of the southern SPI, however, not on the whole study site. Averaging over such a large area can cause different signals in different regions to equal out. The study site does stretch over a large latitudinal band, and we know from literature that the climate and glaciology of NPI and SPI can show different characteristics and patterns. This is taken into account by calibrating the SMB model for both icefields individually, but then ignored throughout the rest of the paper.
Overall, I think by the spatial averaging a lot of valuable information is lost. I advise to conduct a spatiotemporal analysis instead of averaging over the SPI and NPI in order to gain information about the regional variability of climatic control on the SMB in Patagonia.
Minor comments
L1: “Northern and Southern Patagonian Icefields” should be “Northern and Southern Patagonian Icefield”
L56 & 76 & 482: “dryer” should be “drier”
L88: The word “scenario” is strongly associated with climate scenarios, I recommend to reformulate
L93: “assess” to “assesses”
L112-123: The paragraph about the study site is a bit short in my opinion. I would include some brief information about major differences between the two icefields (e.g., SPI many marine terminating glaciers; do we have substantial climatic differences between the two icefields?). A reference to Fig. 2a would make sense here.
L125-131: Which exact variables are taken form RecCMv4?
L133: Why are two different versions used for precipitation and temperature?
L132-140: Both datasets, RefCMv4 and CR2MET, are (at least partly) based on ERA-Interim. I miss a comparison with an independent dataset. What about weather station data, or an independent Reanalysis dataset?
L150: You used the abbreviation ERA-Interim before. Introduce it at the first mentioning, please.
L154: Dot is missing at the end of the sentence.
L158: This is not clear to me: “Lastly, we spatially unweighted averaged the meteorological forcing and the glaciological over the Patagonian Icefields…”
L164: The first “DEM” can be removed.
L183f.: This is not a downscaling of radiation, but simple interpolation.
L199: I would replace the 10800s in the equation by a variable representing the timestep
L201: These are not soil. Rather call it type of surface.
L208: Accurately, the end of summer season would be the 31 March.
L218: Please, use a consistent number of decimal places.
L221-223: The values for c0 are very different between NPI and SPI. Why is this the case?
L227: See comment to L132-140.
L241 & Table1: I recommend using a different abbreviation for the time period here to avoid confusion, as T has been used for temperature before.
Table 1 and following tables: It is common to put the table captions above the respective table.
Table 2: The annual SMB and precipitation value does not exactly add up from winter and summer values. Rounding error?
L286-289: Only mention the significant correlations here: “Among the modeled meteorological variables, the annual SMB is found to have the largest correlation with the annual precipitation (r = 0.69), followed by annual insolation (r = −0.44) (see Table 3). The same order is also evident in winter. The correlation between the SMB and temperature is only significant in summer.”
L332: The correlation seems to be highest especially over the SPI?
L346: “shows” to “show”
Fig. 6b: The grey and white shading is confusing at the first glance, as it seems like there would be two different variables in this plot like it is in panel d. Maybe you can give the shading the same color as the contours to make it clearer.
L368f.: Refer to Fig. 9a here first.
L392ff.: The low correlation with the ENSO and SAM could be due to the spatial averaging over the whole study site. Consider differentiating into regions.
L418-426: Discussion and comparison with other SMB studies in southern Patagonia would support your findings. Similar findings have been found before, e.g., at Grey and Tynall Glacier (Weidemann 2018, https://doi.org/10.3389/feart.2018.00081)
L473 & 490: “SBM” to “SMB”
L476-489: Every paragraph starts with “years of … SMB are characterized by …”. Consider reformulating.
Citation: https://doi.org/10.5194/egusphere-2022-603-RC3 - AC3: 'Reply on RC3', Tomás Carrasco-Escaff, 17 Oct 2022
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2022-603', Anonymous Referee #1, 30 Aug 2022
Review of “Climatic control of the surface mass balance of the Patagonian Icefields” by Carrasco-Escaff et al., submitted to The Cryosphere.
The authors present an interesting study that adds valuable new knowledge to climate and glacier science related to southern South America. The study has been carried out well and is sufficiently documented over large parts of the manuscript. Just the description of the sensitivity analysis is in parts hard to follow and some efforts should be undertaken to improve readability of this section. Apart from this, I have two major objections that prevent me from supporting publication of the article in its present form:
Major comment 1)
The downscaling of solar radiation as it is described in the one sentence provided in L183f has to be questioned. Bilinear interpolation of shortwave radiation on a non-systematically varying surface (like a DEM representing natural terrain) leads to wrong values at the higher-resolution scale. The angle between incoming direct solar radiation and surface slope/aspect (incidence angle) is crucial in determining the right amount of energy reaching the glacier surface. Hence, simply interpolating radiation values from low- to high-resolution grids introduces errors that could easily double or halve solar radiation energy reaching the surface. Regarding diffuse radiation, the skyview factors of the high-resolution grid cells might probably differ considerably from those of low-resolution fields. Taken together, it requires more to downscale solar radiation than just bilinear interpolation.
As spatiotemporal variability of solar geometry can easily be implemented in a downscaling model, the approach needs to be refined by considering incidence angles at each grid cell of the high-resolution topography. Otherwise, the resulting values are simply wrong. Moreover, a validation needs to presented that compares original and downscaled values to in situ measurements (ideally at an on-glacier weather stations). Such a validation must also be presented for T and P, as otherwise it is hard to argue why the RegCM fields can be used for reliable SMB modeling, especially as they show considerable biases to the reference CR2MET climate, which are corrected in a rather simple way only. I’m sure that the team of authors has access to such data even if it might cover only a short period of time.
These validations might also help to overcome the problem of validating the modeled SMB with respect to inter- and intra-annual variability. Assuming that downscaled T, P and R clearly show seasonal variability on a local scale, this would also suggest that the modeled SMB might be reliable in this respect.
Major comment 2)
Climate forcing is analyzed using the SMB integrated over NPI and SPI together. This spatially undifferentiated way of looking at the outcome of this study is a missed opportunity that should be accounted for in a revised and extended version of the study. In its present form the analysis prohibits to get an idea about potential regional variability of forcing mechanisms across Patagonia. I would like to see similar figures to Figs. 6-11 be added to the supplement that show the correlations with only NPI and SPI. Analyzing the differences of these two sets of maps/graphs would give valuable insight into regional variations of climate forcing across Patagonia. This would strengthen the interpretation of the so far presented results which just integrate over NPI and SPI. Sections 3.3-3.5, as well as discussion and conclusion should then be extended accordingly. As we know from the literate that NPI and SPI do not always show the same patterns of glacier change, such an analysis might be of really high value to science – even if it shows that climate forcing mechanisms do not differ significantly for NPI and SPI.
In addition to these comments I have quite some minor comments that also needs some attention of the authors. Based on the two major comments above and the minor comments below, I suggest to return the manuscript for major revision.
Minor comments:
L9: better: ...fields of climate variables from the ERA-Interim…
L40: These positive trends fit to the recent southward shift and strengthening of the southern hemispheric westerly wind belt (e.g. Goyal et al. 2021, doi:10.1029/2020GL090849), which might be of interest here.
L55-57: These moister than average conditions in southern Patagonia have already been suggested to significantly influence SMB (Möller et al. 2007, doi:10.3189/172756407782871530), which should be noted here.
L80: better: …, i.e. the net change of mass at the surface, … “Gain” suggest an increasing mass of ice, but SMB has been positive and negative in the period studied. See Cogley et al. 2011 (Glossary of Glacier Mass Balance) for further details on the related terminology.
L81ff: I see no need to explain glacier mass balance in such detail as the manuscript is written for the cryosphere-centered journal. E.g. basal melting should only be mentioned if it is of interest at the glaciers modeled in the presented study.
L95ff: Braun et al. 2019 and Dussaillant et al. 2019 (both in the manuscript) should also be mentioned here. And it should be discussed that these two remote sensing studies have shown strong mass loss especially over the SPI, which contrasts the positive SMB mentioned before. In its present form the reader gets a picture of increasing ice masses in southern Patagonia, which is wrong.
L129: Why ERA-Interim and not ERA5 which is available for quite a while now?
L134: Also provide reference to Alvarez-Garreton et al. 2018 here, and not only at the end of the paragraph.
L132-140: What makes the CR2MET dataset a reliable reference? I do not question here that it could be used as this, but I would greatly appreciate additional argumentation. It is necessary to outline and explain how well this dataset represents in situ conditions. Moreover, information about shortcomings and especially inaccuracies of the dataset are needed to be able to judge about its reliability. And finally (maybe most important) why are the RegCM fields created and used when CR2MET already exists? What is the advantage of RegCM over CR2MET and does this advantage justify the introduction of additional uncertainty (by comparing it to CR2MET before usage)?
L147: better: “… of world-wide glacier extent at the beginning…”, as “extension” implies a process of increase rather than a static condition
L158: not clear what is meant here: “Lastly, we spatially unweighted averaged the meteorological forcing…”
L159: better: “Only grid points within…” (omit “Note that”)
L192: provide reference for this representation of the fraction of solid precipitation
L209ff: It would be interesting to get some values on the distribution of snow/firn after the spin-up time: Give average numbers for snow-/firnline altitudes across the study area and discuss potential spatial variations in case they exist. Give reference to other studies which derived snowline altitudes in Patagonia and shortly compare your results to these findings.
L231-235: This is a really nice idea. However, I strongly request that also information about the bias in SMB compared to the reference SMB is somehow incorporated in the Taylor diagram (e.g. by scaled sizes or color-scales of the points shown). The so far given information about correlation and standard deviation only give insight into how well the variability is represented, but do not tell anything about resulting biases.
L239-249: This is an interesting approach, but more information is needed here. First, give reference to studies that introduced or at least support your idea. Second, give more details on how you determined the variability in the dataset and how you subsequently removed it. Also here, a quantification of biases is needed in addition to the measures of variability.
Fig. 5: I suggest to add a thin black line representing a zero SMB in the upper panel of the figure. This would increase readability and make positive and negative SMB years more easily distinguishable.
Table 2: Add information about the period represented by the given numbers to the caption.
L287: The fact that annual insolation shows a higher correlation to SMB than annual temperature further supports my initial request regarding a refined handling of solar radiation during downscaling.
L304ff: Isn’t that a necessary result of the over-simplified radiation downscaling that has been applied? I mean, how can a local-scale control over the SMB can be present when the applied downscaling is not able to produce the requited local-scale variability? (see my initial major comment) This analysis/interpretation must be redone after the radiation downscaling has been improved.
L307-318: It now entirely clear what was done here. A linear regression results in intercept and slope of a regression line, which are both important for interpretation. However, this full information is missing in Table 4 and has to be added. It must also be included in the following discussion.
L325ff: Why is solar radiation not considered here?
Figs. 6b/7b: I recommend not to use red/green colors for the isolines as these colors are hard to differentiate for a lot of color-blind people.
L410ff: It would greatly strengthen the findings of the study if comparisons to other long-term SMB time series at other Patagonian glaciers would be given. E.g. Möller & Schneider 2008 (doi:10.3189/172756408784700626) present a modeled SMB time series for Gran Campo Nevado ice cap south of the SPI. This time series e.g. shows the same strongly positive anomalies of SMB in 1990 and 1995, which supports the presented findings for SPI by showing that they fit nicely into the picture presented by other studies. Further south (e.g. Tierra del Fuego) other SMB pattern prevail (e.g. Buttstädt et al. 2009, doi:10.5194/adgeo-22-117-2009), suggesting a southward limitation of the regional pattern.
L418: Doesn’t this contradict the results that you presented before (see my comments on L287 and L304ff)? This should be clarified either here and/or above.
L418-426: This paragraph would benefit from some references to either figures or tables.
L456ff: References to other studies dealing with this or comparable issues would support your speculation and should be added and discussed shortly.
L474: This thought has not come to my mind until now: Is there any significant interannual variability in solar radiation? Or is it largely time-invariant? I’m asking because of the frequent presence of clouds in Patagonia. If there is no significant interannual variability, it would be a necessary consequence that SMB variations show almost not dependence on it. This needs to be analyzed (and outlined in the results section) before giving this broad statement, in order to potentially put it into the right context.
L490: “SBM” needs to be corrected to “SMB”
Citation: https://doi.org/10.5194/egusphere-2022-603-RC1 - AC1: 'Reply on RC1', Tomás Carrasco-Escaff, 17 Oct 2022
-
RC2: 'Comment on egusphere-2022-603', Anonymous Referee #2, 31 Aug 2022
Summary: this paper describes the climatic controls on the surface mass balance (SMB) of the North and South Patagonian Icefields (NPI, SPI). This is achieved by estimating the annual and seasonal SMB with a simple snow, firn and ice accumulation and ablation model, subsequently regressing the SMB-anomalies time series to a suite of local, regional and climate indices. Results indicate that winter precipitation and summer temperature anomalies are the main drivers of SMB interannual variability. Also, the authors find that a pressure anomaly over the Drake’s passage (the Drake low) is the dominant feature related to SMB departures, seemingly driving increased westerly winds and cooler conditions off the coast of Patagonia. No significant correlation was found between the SMB and major climate indices such as ENSO, which confirms previous work published in the area.
General comments: this is a well written paper, and is a nice contribution to the understanding of the NPI and SPI present-day behavior. The authors have taken preemptive actions to prevent the inevitable modeling uncertainties from affecting their conclusions, by focusing on correlations/anomalies only and by ensuring that potential biases in the meteorological forcings of the SMB model don’t result in major changes in the year to year variability, measured through correlation and standard deviation of the time series. The organization of the manuscript is very intuitive and the use of English language is appropriate but for a few minor issues. Because the analysis rests so strongly on the simulated mass balance, the manuscript should devote a bit more space to discussing the calibration of the four main parameters of the model, namely the threshold at which precipitation falls as snow (here set as 2°C), and the ablation parameters (albedo, c_0 and c_1). The sensitivity of the model to these parameters should in turn influence the interplay between precipitation and temperature during the accumulation season, and the relative influence of radiation and temperature during the ablation season. It may be that the main conclusions don’t change with respect to what is shown in the current version, but so far the paper seems to gloss over this topic in a manner too succinct.
Specific comments:
L132: It is not clear to me what the verification of RegCMv4 against CR2MET intends to achieve. There are clear biases shown in Fig2, which could result from several factors. Because you have threshold term in accumulation that depends on T, this bias in temperature could have compounded effects on the simulated SMB correlations. Do you do anything after verifying the two products against each other?
L201: snow, firn and ice should not be called “soil”. Please use something like “land cover”.
L210: in modeling parlance, “true” has a very specific meaning. Please revise.
L218: See general comments regarding the detail that is needed about the SMB model calibration process. Also, why do you compare the 2000-15 simulation with the 2000-19 Minowa estimates? Is it not possible to compare a common period?
L229: These biases could also result from inadequacies in the CR2MET product. In particular, if it is station-based, previous research has shown that meteorological station data in Patagonia is unreliable, particularly precipitation.
L231: a similar analysis could be performed by perturbing some of the model parameters (see general comments).
L260: If I understand correctly, AAO was calculated only for the 1979-2000 period? But SMB is available until 2015? Maybe it’d be useful to have a summary table with all datasets used, indicating time window, time-step, and citation.
L272: please reword to remind the reader that all these numerical quantities are estimates from your model. Also, the fact that annual SMB is positive means that for the ice fields to be in equilibrium (or decreasing in mass, as the literature suggests) then calving should account for the excess mass. Is that right? Also: there appears to be a slight increasing trend in the simulated SMB? Could you comment on this?
L298: How do you interpret the fact that although insolation shows the second-highest correlation with annual SMB (line 287), then the local-scale control indicates exactly the opposite? Maybe I’m missing something, but these two results seem inconsistent. Please clarify. Is this result sensitive to assumptions regarding the seasonal evolution of snow, firn and ice albedo? Nevertheless, it is expected that, unlike glaciers in mediterranean regions, solar radiation should have a minor role compared to temperature in Patagonia. High relative humidity and high very persistent cloud cover are coherent with this result.
L406: a small detail: probably “good” is not an appropriate adjective for describing correlation.
L446: suggest replacing “maintains” with “remains”.
Citation: https://doi.org/10.5194/egusphere-2022-603-RC2 - AC2: 'Reply on RC2', Tomás Carrasco-Escaff, 17 Oct 2022
-
RC3: 'Comment on egusphere-2022-603', Anonymous Referee #3, 13 Sep 2022
General comments
The study analyzes the control of the present-day climate on the surface mass balance of the Patagonian Icefields. The main goals of the study are clearly formulated, and the study is well structured and written over largest parts. In the Discussion section, a stronger comparison with and discussion of results of other SMB studies in the region could strengthen the findings. Overall, the study adds valuable knowledge to the understanding of the interaction between climate and glacier mass balance in the southern Andes. I have one major comment which needs addressing and revision before publishing the article, together with several minor comments.
Major comment
The main limitation of this study is the fact that there is no spatial analysis for the correlation of SMB with large-scale indices and climate. It is possible that e.g. SAM does have an important impact on the SMB of the southern SPI, however, not on the whole study site. Averaging over such a large area can cause different signals in different regions to equal out. The study site does stretch over a large latitudinal band, and we know from literature that the climate and glaciology of NPI and SPI can show different characteristics and patterns. This is taken into account by calibrating the SMB model for both icefields individually, but then ignored throughout the rest of the paper.
Overall, I think by the spatial averaging a lot of valuable information is lost. I advise to conduct a spatiotemporal analysis instead of averaging over the SPI and NPI in order to gain information about the regional variability of climatic control on the SMB in Patagonia.
Minor comments
L1: “Northern and Southern Patagonian Icefields” should be “Northern and Southern Patagonian Icefield”
L56 & 76 & 482: “dryer” should be “drier”
L88: The word “scenario” is strongly associated with climate scenarios, I recommend to reformulate
L93: “assess” to “assesses”
L112-123: The paragraph about the study site is a bit short in my opinion. I would include some brief information about major differences between the two icefields (e.g., SPI many marine terminating glaciers; do we have substantial climatic differences between the two icefields?). A reference to Fig. 2a would make sense here.
L125-131: Which exact variables are taken form RecCMv4?
L133: Why are two different versions used for precipitation and temperature?
L132-140: Both datasets, RefCMv4 and CR2MET, are (at least partly) based on ERA-Interim. I miss a comparison with an independent dataset. What about weather station data, or an independent Reanalysis dataset?
L150: You used the abbreviation ERA-Interim before. Introduce it at the first mentioning, please.
L154: Dot is missing at the end of the sentence.
L158: This is not clear to me: “Lastly, we spatially unweighted averaged the meteorological forcing and the glaciological over the Patagonian Icefields…”
L164: The first “DEM” can be removed.
L183f.: This is not a downscaling of radiation, but simple interpolation.
L199: I would replace the 10800s in the equation by a variable representing the timestep
L201: These are not soil. Rather call it type of surface.
L208: Accurately, the end of summer season would be the 31 March.
L218: Please, use a consistent number of decimal places.
L221-223: The values for c0 are very different between NPI and SPI. Why is this the case?
L227: See comment to L132-140.
L241 & Table1: I recommend using a different abbreviation for the time period here to avoid confusion, as T has been used for temperature before.
Table 1 and following tables: It is common to put the table captions above the respective table.
Table 2: The annual SMB and precipitation value does not exactly add up from winter and summer values. Rounding error?
L286-289: Only mention the significant correlations here: “Among the modeled meteorological variables, the annual SMB is found to have the largest correlation with the annual precipitation (r = 0.69), followed by annual insolation (r = −0.44) (see Table 3). The same order is also evident in winter. The correlation between the SMB and temperature is only significant in summer.”
L332: The correlation seems to be highest especially over the SPI?
L346: “shows” to “show”
Fig. 6b: The grey and white shading is confusing at the first glance, as it seems like there would be two different variables in this plot like it is in panel d. Maybe you can give the shading the same color as the contours to make it clearer.
L368f.: Refer to Fig. 9a here first.
L392ff.: The low correlation with the ENSO and SAM could be due to the spatial averaging over the whole study site. Consider differentiating into regions.
L418-426: Discussion and comparison with other SMB studies in southern Patagonia would support your findings. Similar findings have been found before, e.g., at Grey and Tynall Glacier (Weidemann 2018, https://doi.org/10.3389/feart.2018.00081)
L473 & 490: “SBM” to “SMB”
L476-489: Every paragraph starts with “years of … SMB are characterized by …”. Consider reformulating.
Citation: https://doi.org/10.5194/egusphere-2022-603-RC3 - AC3: 'Reply on RC3', Tomás Carrasco-Escaff, 17 Oct 2022
Peer review completion
Journal article(s) based on this preprint
Tomás Carrasco-Escaff et al.
Tomás Carrasco-Escaff et al.
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 444 | 201 | 20 | 665 | 46 | 6 | 6 |
- HTML: 444
- PDF: 201
- XML: 20
- Total: 665
- Supplement: 46
- BibTeX: 6
- EndNote: 6
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(12214 KB) - Metadata XML
-
Supplement
(4662 KB) - BibTeX
- EndNote
- Final revised paper