The importance of model horizontal resolution for improved estimation of snow water equivalent in a mountainous region of western Canada

Sabetghadam, Samaneh; Fletcher, Christopher G.; Erler, Andre

doi:https://doi.org/10.5194/hess-29-887-2025

Articles | Volume 29, issue 4

https://doi.org/10.5194/hess-29-887-2025

© Author(s) 2025. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/hess-29-887-2025

© Author(s) 2025. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 29, issue 4

Research article

|

20 Feb 2025

Research article |

| 20 Feb 2025

The importance of model horizontal resolution for improved estimation of snow water equivalent in a mountainous region of western Canada

Samaneh Sabetghadam, Christopher G. Fletcher, and Andre Erler

Download

Final revised paper (published on 20 Feb 2025)
Preprint (discussion started on 06 Mar 2024)

Interactive discussion

Status: closed

RC1:
'Comment on egusphere-2024-42', Anonymous Referee #1, 19 Mar 2024

The comment was uploaded in the form of a supplement

Citation: https://doi.org/10.5194/egusphere-2024-42-RC1
- AC1: 'Reply on RC1', Samaneh Sabetghadam, 10 Jun 2024
  
  The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2024-42/egusphere-2024-42-AC1-supplement.pdf
  
  Citation: https://doi.org/10.5194/egusphere-2024-42-AC1
RC2:
'Comment on egusphere-2024-42', Anonymous Referee #2, 14 May 2024

In general it is hard to read the figures — the authors should consider saving them at a higher resolution and/or increasing font sizes. This is making it harder to interpret the figures.

"Figure 2. Temporal variation of SWE, from CanSWE data and WRF model resolution over the SSRB region. The SWEdata are aggregated for all stations inside the innermost domain. The spikes in the graph correspond to snowfall events."
Do the spikes imply that the snow accumulated and then melts, or that this is a data artifact? The authors should state what they think is actually causing the spikes. It seems unusual to have the spikes consistent across all of the CanSWE observations.
"Figure 6. Simulated mean SWE (kg/m2)"
kg/m² should be properly formatted
"is is because the actual topography and associated elevation will typically be much more variable over an area of 81 km2 than over 1 km..."
Same here — correctly format km².
"The large bias in the coarsest resolution also may be due to the incapability of the WRF-9km to simulate the processes that are responsible for snow deposition and erosion in mountainous areas (Mott et al., 2018; Raparelli et al. 2021)."
It is not clear what "snow erosion" is in this context. Maybe "snow redistribution"? The latter is the more commonly used term.

Citation: https://doi.org/10.5194/egusphere-2024-42-RC2
- AC2: 'Reply on RC2', Samaneh Sabetghadam, 10 Jun 2024
  
  The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2024-42/egusphere-2024-42-AC2-supplement.pdf
  
  Citation: https://doi.org/10.5194/egusphere-2024-42-AC2

Peer review completion

AR: Author's response | RR: Referee report | ED: Editor decision | EF: Editorial file upload

ED: Reconsider after major revisions (further review by editor and referees) (17 Jun 2024) by Hongkai Gao

AR by Samaneh Sabetghadam on behalf of the Authors (18 Jun 2024) Author's response Author's tracked changes

EF by Sarah Buchmann (21 Jun 2024) Manuscript

ED: Referee Nomination & Report Request started (23 Jun 2024) by Hongkai Gao

RR by Anonymous Referee #1 (07 Aug 2024)

RR by Anonymous Referee #3 (02 Dec 2024)

Suggestions for revision or reasons for rejection

The study evaluates the Weather Research and Forecasting (WRF) model at three different resolutions (9km, 3km, 1km) for estimating Snow Water Equivalent (SWE) in Western Canada's South Saskatchewan River Basin (SSRB). It uses CanSWE dataset for validation during the 2017-2018 water year, Employs ERA5 reanalysis data for boundary conditions

Major Findings:
• Higher resolution (1km and 3km) WRF simulations significantly improve SWE estimates compared to coarser resolution (9km)
• The 9km resolution consistently underestimates SWE by about 31%
• Peak SWE timing is well captured across all resolutions, occurring in late April
• Strong correlation between elevation bias and SWE bias, particularly at coarser resolutions
Scientific Merit:
• The methodology is robust, using established models and validation techniques
• Results align with theoretical expectations about resolution impacts on mountainous precipitation
• Good use of statistical analysis to quantify model performance
Areas for Improvement:
From a climate modeling perspective:
• The study could benefit from analyzing more water years to ensure results are not year-specific
• Further investigation of physical mechanisms behind resolution-dependent biases would strengthen the findings
• Discussion of computational costs vs. benefits of higher resolution would be valuable

Certain sections could be rewritten for better clarity, here some examples:

1 - From Abstract:
Original: "All WRF simulations tend to underestimate annual SWE, with largest biases (up to 58 kg/m2, i.e. relatively 24%) found at higher elevations and in simulations at coarser horizontal resolution."

Can be improved: "All WRF simulations underestimated the annual SWE. The largest errors occurred in two conditions: at higher elevations and when using coarser horizontal resolution. These biases reached up to 58 kg/m2 (24% relative error)."

2 - From Introduction:
Original: "Spatiotemporal distribution of SWE, particularly within northern latitudes and higher elevations, shows the extent of spring and summer runoff (Barnet et al., 2005; King et al., 2020)."

Can be improved: "The distribution of SWE across space and time is especially important in northern regions and at high elevations. This distribution determines how much water will be available during spring and summer runoff periods (Barnet et al., 2005; King et al., 2020)."

3 - From Results:
Original: "Examining the role of these two forcings on simulated SWE at the three resolutions, we find very close agreement in temperature (Fig.3a), but a systematic low bias in accumulated precipitation at WRF9K (Fig.3b), indicating that lower total precipitation is the most likely cause of the SWE bias at the lowest resolution."

Can be improved: "We examined how temperature and precipitation affect SWE simulation at different resolutions. Temperature values were very similar across all three resolutions (Fig.3a). However, the WRF9K consistently showed lower precipitation values than expected (Fig.3b). This suggests that the underestimation of precipitation, rather than temperature differences, is the main cause of SWE bias in the lowest resolution model.

4 - From Discussion:
Original: "The large bias in the coarsest resolution also may be due to the incapability of the WRF9K to simulate the processes that are responsible for snow deposition and redistribution in mountainous areas that are characterized by heterogeneous snow distribution."

Can be improved: "The WRF9K model shows large errors primarily because of its resolution limitations. At 9km resolution, the model cannot accurately capture two key processes in mountainous areas: snow deposition and snow redistribution. These processes are especially important in mountains because snow distribution varies greatly over short distances."

Hide

ED: Publish subject to minor revisions (review by editor) (17 Dec 2024) by Hongkai Gao

AR by Samaneh Sabetghadam on behalf of the Authors (19 Dec 2024) Author's response Author's tracked changes Manuscript

ED: Publish as is (29 Dec 2024) by Hongkai Gao

AR by Samaneh Sabetghadam on behalf of the Authors (01 Jan 2025) Manuscript

Short summary

Snow water equivalent (SWE) is an environmental variable that represents the amount of liquid water if all the snow cover melted. This study evaluates the potential of the Weather Research and Forecasting (WRF) model to estimate the daily values of SWE over the mountainous South Saskatchewan River Basin in Canada. Results show that high-resolution WRF simulations can provide reliable SWE values as an accurate input for hydrologic modeling over a sparsely monitored mountainous catchment.