Sensitivity of snow models to the accuracy of meteorological  forcings in mountain environments

Terzago, Silvia; Andreoli, Valentina; Arduini, Gabriele; Balsamo, Gianpaolo; Campo, Lorenzo; Cassardo, Claudio; Cremonese, Edoardo; Dolia, Daniele; Gabellani, Simone; von Hardenberg, Jost; Morra di Cella, Umberto; Palazzi, Elisa; Piazzi, Gaia; Pogliotti, Paolo; Provenzale, Antonello

doi:https://doi.org/10.5194/hess-24-4061-2020

Articles | Volume 24, issue 8

https://doi.org/10.5194/hess-24-4061-2020

© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/hess-24-4061-2020

© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 24, issue 8

Research article

|

21 Aug 2020

Research article |

| 21 Aug 2020

Sensitivity of snow models to the accuracy of meteorological forcings in mountain environments

Silvia Terzago, Valentina Andreoli, Gabriele Arduini, Gianpaolo Balsamo, Lorenzo Campo, Claudio Cassardo, Edoardo Cremonese, Daniele Dolia, Simone Gabellani, Jost von Hardenberg, Umberto Morra di Cella, Elisa Palazzi, Gaia Piazzi, Paolo Pogliotti, and Antonello Provenzale

Download

Final revised paper (published on 21 Aug 2020)
Preprint (discussion started on 25 Nov 2019)

Interactive discussion

Status: closed

AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment

- Printer-friendly version

- Supplement

RC1: 'Referee comments', Anonymous Referee #1, 20 Dec 2019
- AC1: 'First reply to Reviewer #1', Silvia Terzago, 20 Dec 2019
- AC2: 'Reply to Reviewer #1', Silvia Terzago, 21 Feb 2020
RC2: 'Review of Terzago et al.', Richard L.H. Essery, 27 Dec 2019
- AC3: 'Reply to Reviewer #2 Richard L.H. Essery', Silvia Terzago, 21 Feb 2020

Peer-review completion

AR: Author's response | RR: Referee report | ED: Editor decision

ED: Reconsider after major revisions (further review by editor and referees) (02 Mar 2020) by Daniel Viviroli

AR by Silvia Terzago on behalf of the Authors (06 May 2020) Author's response Manuscript

ED: Referee Nomination & Report Request started (15 May 2020) by Daniel Viviroli

RR by Richard L.H. Essery (31 May 2020)

Suggestions for revision or reasons for rejection

Some technical corrections are required:

Figures 1 and 4
I does not make sense for the grid lines to extend beyond the axes, and it is not clear what the spacing of the vertical (date) grid lines is supposed to represent.

Figure 2
The legend giving the model names overlaps the “correlation” axis label.

Figure 5
More space is required between the axis labels and numbers.

Figure 6
The grey backgrounds have no purpose and should be removed. Use a superscript for m2.

page 25
Three hours is not the typical temporal resolution of climate and weather forecast model simulations; this is just the resolution of archived outputs.

Figure A1
Label the months in English

Appendix D
Why would ERA-Interim inherently take orographic effects into account and ERA5 not?

I also suggest corrections to the text:

page 1, line 12
“GLDAS, ERA5 and ERA-Interim reanalyses”

page 2, line 1
“if the incoming shortwave radiation is not properly represented” would be better
line 2
“reanalyses are found to be adequate forcings” would be better
line 23
“in Russian and Swedish snow-dominated catchments”
line 25
“the reasons for model scatter”

page 3, line 15
“and avalanche forecasting,”
line 21
“but simple snow models with an appropriate parameterization of albedo dynamics also guaranteed”
line 26
“Calibration allowed reduction of root mean square error”

page 4, line 11
“for snow modelling at the catchment scale”
line 14
“uncertainty of snow simulations”

page 5, line 6
“and finally Sections 6 and 7”
line 18
“The model also takes into account snow metamorphism”
line 24
“allowing representation of the interception of rainfall”
line 28
“a single homogenous layer placed on the land surface”

page 6, line 5
“only a few input variables”
line 7
“obtained by combining”

page 7, line 3
“most precipitation falls as snow”
line 11
“higher confidence in the reliability”
line 15
“is assessed and discussed in Appendix A.”
line 18
“The Torgnon station also provides snow-related variables”
line 23
“Campbell CS725 sensor, that passively measures” – delete the comma
line 25
“The measurement is performed”

page 8, line 19
“ERA5 also has”
line 25
“This run allows testing the accuracy of the models”
line 29
“when obstructed by snow”

page 10, line 4
“more details are provided in Appendix C”
line 11
“reanalysis data had to be rearranged”
line 24
“to what extent”
line 27
“In the ERAI-LR experiment”
line 28
“This correction results in a cooling”
line 32
“differences in the model outputs should be mainly due to the internal model characteristics” ?

page 11, line 7
“allows reduction of the range”
line 9
“depend mainly on”

page 12, line 7
“reproduces all three variables with high scores”

page 13, line 14
“Figure 2e allows investigation of the variability”
line 22
“mineral dust from the Sahara”

page 14, line 2
“In these conditions”
line 9
“which keeps the date of ablation”

page 17, line 18
“in the case of total precipitation”
line 28
“reflected in an increase”

page 19, line 1
“RMSE values associated with”

page 20, line 5
“and GLDAS, ERA5 and ERA-Interim reanalyses”
line 12
“similar RMSE values to the CTL runs”

page 21, line 11
“error in snow depth”
line 13
“the error in the total precipitation”
line 18
“but with thicker snowpacks”
line 19
“consistent with the behavior found in the CTL run”
line 21
“leading to overestimation”

page 22, line 8
“the snowfall bias increases”

page 23, line 4
“almost removes the snowfall bias”
line 6
“this approach requires knowing”
line 7
“reduces the RMSE in snow water equivalent”
line 12
“on average”
line 13
“the error in snow density (lower correlation with available observations) and thus the error in snow depth simulations”
line 18
“still need to be explored”

page 25, line 3
“using alternative radiation data”
line 5
“obtained replacing measured data”
line 9
“RMSE in snow depth”
line 20
“allows improving the snow simulations and model skills comparable to”
line 22
“the bias in the incoming shortwave radiation flux”
line 24
“improved SWIN input allows surprisingly good performances”
line 32
“the latter can also be”
line 33
“consequently also adjusting”

page 26, line 8
“in the absence of in-situ observations. In our experiment the models’ RMSE values”
line 12
“a precipitation bias that varies strongly from season to season”
line 15
“The present analysis allows straightforward evaluation of the performances of each model”
line 28
“error in snow depth. These errors offset each other, with the effect that the RMSE in snow depth”
line 29
“GEOTOP performances are the worst owing to the positive temperature bias”

page 27, line 9
“performed at a single mountain site”
line 13
“obtained at other sites”
line 19
“provides guidance”
line 33
“lower RMSE values in snow depth”

page 28, line 19
“the possibility of using lower frequency data (up to 3 hours) without losing accuracy in the snow output”

page 31, line 6
“compared with both observations and the CTL run”
line 10
“we decided to employ the original precipitation measurements”

Figure B1 caption
“represents the extent of the digital elevation model”

page 35, line 14
“suffer from underestimation of precipitation”

page 36, line 9
“biases at the Torgnon site are spatially consistent”
line 13
“All authors participated”
line 14
“ST analysed the simulations”

Hide

ED: Publish subject to minor revisions (review by editor) (01 Jun 2020) by Daniel Viviroli

AR by Silvia Terzago on behalf of the Authors (11 Jun 2020) Author's response Manuscript

ED: Publish as is (30 Jun 2020) by Daniel Viviroli

AR by Silvia Terzago on behalf of the Authors (03 Jul 2020) Manuscript

Short summary

In mountain areas high-quality meteorological data to drive snow models are rarely available, so coarse-resolution data from spatial interpolation of the available in situ measurements or reanalyses are typically employed. We perform 12 experiments using six snow models with different degrees of complexity to show the impact of the accuracy of the forcing on snow depth and snow water equivalent simulations at the Alpine site of Torgnon, discussing the results in relation to the model complexity.