Earth System Model-Snow Model Intercomparison Project (ESM-SnowMIP)

Snow is an essential component of the Earth System. As both a permanent and an intermittent component of the cryosphere, it interacts with the atmosphere and the surfaces it covers (land, ice, sea ice) and it is one of the principal sources of feedbacks within the climate system. As an outcome of the Tromsø 2013 Workshop the WCRP/CliC Initiative for a ESM-SnowMIP intercomparison programme is a contribution to the WCRP Grand Challenge Melting Ice & Global Consequences. The current strengths and weaknesses of snow models used in ESM and GCM must be assessed in order to provide guidelines for their improvement.

Why is a large-scale integrated study of snow in the climate system timely?

Col de Porte snow and meteorology field site, French Alps. - Y. Lejeune / Météo-France - CNRS The most well-known positive feedback stems from the albedo contrast between snow-covered and snow-free surfaces. It can lead to a local warming and critically determine the conditions for persistence or decay of permafrost. Undergoing quick changes, snow is also a prominent and powerful indicator of climate change. Relative changes in seasonal snow can be as large as or even stronger than those in sea-ice extent. Last but not least, snow is a critical determinant for water resources in many regions of the Earth.

In spite of this, there are still fundamental knowledge gaps concerning the physical properties of snow and their evolution in time in both seasonal and permanent snow packs, past, present and future trends in snow mass and distribution, and the effects of snow in the Earth System. The main scientific questions are:

1. How can we improve our current knowledge and understanding of the temporal dynamics and physical properties of snow as a component of the coupled climate system?
2. How can we improve our understanding of the role of snow as an active component of the global climate system?

There is a clear need to better tie together observational data, remote sensing, and modelling using novel approaches available through WCRP. This is something that has not been tackled before within a large international framework such as under CliC. Field experiments, dataset generation, model development, and process studies need to be done together at a meaningful range of time and spatial scales, and could be complemented by a multi-scale observing and modelling experiment.

ESM-SnowMIP: Building on a strong legacy

As a first part of this Snow in the Earth System Study, a targeted action on the intercomparison and improvement of snow modules of large-scale climate models has been initiated as a part of the Melting Ice & Global Consequences Grand Challenge. This project is called ESM-SnowMIP (for « Earth System Model snow module intercomparison/improvement project »). As its name already says, it is specifically designed to evaluate and improve snow modules of large global climate or Earth System models that, on one hand, can be used to evaluate the role of snow in the Earth System and, on the other hand, require a correct representation of snow and its effects on the climate system even for applications not specifically focussed on snow. Indeed, the amplitude of snow‐related feedbacks, the effect of snow on other components of the cryosphere, its effect on climate predictability (seasonal or longer term) are not well quantified.
Over the past decade, much progress has been made on the representation of snow in dedicated models, in particular through the SnowMIP 1 & 2 projects (e.g., http://www.geos.ed.ac.uk/~ressery/SnowMIP2.html). However, this progress has not yet been transposed to snow modules of the larger Earth System models. This lead to the following, still valid conclusion of the GSWP2 project: « Generally there is mediocre agreement among the models for most of the snow-related variables, suggesting a potential area of continuing weakness in global land surface schemes » (Dirmayer et al., 2006). It is time now to close this gap.

The goal of ESM-SnowMIP is therefore to make progress on the representation of snow in ESMs, and conduct systematic model studies and data analysis for a better understanding of the role of snow in the global climate system.
ESM-SnowMIP is an intercomparison programme designed to quantify the state of the art of the representation of snow in ESM and RCMs and identify ways to move forward (building on the experience of SnowMIP and extending it specifically to ESM and RCM land surface modules). The intercomparison involves dedicated detailed snow models to gauge required complexity and necessary processes to be represented in ESMs, and an assessment of snow-related processes and feedbacks in CMIP5/6.
This action could, and actually should, be part of a broadescope higlatitude surface processes model intercomparison addressing permafrost physics and biogeochemistry, snow (on sea ice, land ice and on land), vegetation dynamics, and their interactions in ESMs and RCMs. As part of the Melting Ice & Global Consequences Grand Challenge action, it is closely linked to the GEWEX-GLASS Global Soil Wetness Project Phase 3 (GSWP-3: http://www.wcrp-climate.org/index.php/modelling-wgcm-mip-catalogue/modelling-wgcm-mips-2/257-modelling-wgcm-catalogue-gswp).

Design of ESM-SnowMIP

The following actions are planned within ESM-SnowMIP:

• Identification of remotely sensed and site snow-related datasets for model evaluation and improvement.
• Evaluation of snow cover simulations over continents and ice-sheets in CMIP5 Historical scenarios to describe the current state of snow representation in ESMs
• Stand-alone snow simulations driven with global/regional gridded meteorological data sets to identify the differences due to the models
• Stand-alone simulations with prescribed soil/vegetation types to identify the effects due to the representation of interactions between snow, soil and vegetation
• Stand-alone snow simulations driven with data sets representative of climate scenarios to assess the intrinsic sensitivity of snow models to changed climate conditions
• ESM / GCM simulations with prescribed and free snow conditions to identify snow feedbacks on the atmosphere
• ESM /GCM simulations nudged towards reanalyses to identify snow/atmosphere feedbacks
• Recommendations to CMIP6 for defining snow variables available via ESG
• Local simulations over a few selected supersites (i.e. representative of polar, alpine, continental, maritime conditions) in order to interpret some of the results
• Simulation of soil temperature and/or thermally active depth to assess the snow/soil coupling in ESM / GSM
• At a later stage, a possible extension to snow models over sea ice (i.e. optional tier or second phase of the project)

The ESM-SnowMIP Simulation Protocol describes the ESM-SnowMIP numerical experiments proposed as complements to the simulations proposed in LS3MIP.

Descriptions of the reference sites and detailed instructions for the reference site simulations can be found here.

ESM-SnowMIP is tightly linked to LS3MIP. LS3MIP is described in the CMIP6 special issue of the journal Geoscientific Model Development:

van den Hurk, B., Kim, H., Krinner, G., Seneviratne, S. I., Derksen, C., Oki, T., Douville, H., Colin, J., Ducharne, A., Cheruy, F., Viovy, N., Puma, M., Wada, Y., Li, W., Jia, B., Alessandri, A., Lawrence, D., Weedon, G. P., Ellis, R., Hagemann, S., Mao, J., Flanner, M. G., Zampieri, M., Law, R., and Sheffield, J. : The Land Surface, Snow and Soil moisture Model Intercomparison Program (LS3MIP) : aims, set-up and expected outcome, Geosci. Model Dev., 9, 2809-2832, doi:10.5194/gmd-9-2809-2016, 2016.

Call for Participation - ESM-SnowMIP

ESM-SnowMIP is a new initiative that aims at evaluating and improving the representation of snow primarily (but not exclusively) in Earth System Models and at better quantifying snow-related climate feedbacks. It consists of a suite of coordinated experiments at the site level and on the global scale, plus numerical experiments in the Earth System Model framework.

ESM-SnowMIP is part of the WCRP Grand Challenge "Melting Ice & Global Consequences" and is designed as a follow-up project to previous SnowMIP initiatives and complements, by a specific focus on snow, the LS3MIP (Land Surface, Snow, and Soil moisture MIP) subproject of CMIP6 that is designed to evaluate land surface modules of Earth System Models and to quantify land-related feedbacks. Compared to previous SnowMIP editions, a new and enlarged set of sites for model evaluation and testing is used.

We solicit participation in ESM-SnowMIP by:

• Earth System Modeling groups, particularly those that participate to the LS3MIP subproject of CMIP6;
• Snow and general land surface modelers (both site- and large scale).

At this stage, our aim is to obtain expressions of interest from the scientific community and receive feedback to the initial ESM-SnowMIP simulation protocol available here.

We anticipate that, besides progress due to coordinated model testing and benchmarking, the common participation by Earth System Modeling groups and specialized snow scientists will generate scientific momentum that will allow substantial knowledge transfer to the Earth System Modeling community, and thus allow for significant improvement of the representation of snow in global climate models.

The general timeline of ESM-SnowMIP aims to initiate site- and global-scale offline simulations in 2016, while coupled model simulations (complementing numerical experiments carried out in the LS3MIP framework) will start in 2018 after CMIP6.

The ESM-SnowMIP steering committee members are: Gerhard Krinner, Chris Derksen, Richard Essery, Stefan Hagemann, Alex Hall, Andrew Slater, Matthew Sturm and Helmut Rott. Please send your expressions of interest to Gerhard Krinner and Chris Derksen.

Related observational activities

Snow extent, water equivalent and physical properties are highly variable in space and time, and knowledge of this variability is of paramount importance to better understand the role snow plays in the global climate system. Within the Melting Ice & Global Consequences Grand Challenge, several efforts will be undertaken to produce better global-scale datasets of snow cover, snow water equivalent, and snow density by drawing together local and remote sensing observations. Good integration of these efforts with ESM-SnowMIP modeling activities is ensured by the presence of several snow observation experts in the ESM-SnowMIP steering committee.
This involves work on identifying existing snow-related observation supersites with a representative variety of regimes, in links with the supersite-related action of the WMO Global Cryosphere Watch (GCW): ice sheets (possibly Dome C), alpine type (possibly Col de Porte), Siberia, sea ice (MOSAIC platform), permafrost-related (possibly Barrow), and developing common long term observation strategies and protocols. These supersites are designed to host comprehensive observations of physical and chemical effects of snow in order to improve scientific process understanding and for the development and evaluation of a new generation of dedicated models and snow models in ESMs. This can start from a small number of sites.
We further envisage a workshop/white paper designed to develop new strategies to exploit the wide variety of snow-related ground-based observations, airborne and space-based remote sensing measurementsfor a comprehensive evaluation of GCM-designed snow models and GCM snow outputs (http://www.climate-cryosphere.org/news/clic-news/855-snowpex2014). This will be done in connection with IASC and WMO GCW.

Reference sites: Site description and instructions for simulations

ESM-SnowMIP workshop in San Francisco, Dec 10, 2016 (prior to AGU)

ESM-SnowMIP held its first workshop in Fort Mason (map available here), San Francisco, on Dec 10, 2016 (Saturday prior to AGU). Topics of the workshop were:

• Presentation and discussion of first preliminary site simulation results
• Analysis strategies for site and global-scale simulations, identification of leaders of these analyses
• Planning of global-scale offline and coupled ESM simulations
• Links to CMIP6 sub-projects: LS3MIP (land-surface, snow and soil humidity) and ISMIP6 (ice sheet modeling)

Many modeling groups and snow observation specialists contributing to ESM-SnowMIP were able to participate in the workshop. The agenda of the 2016 meeting is available here.  The list of action items emanating from the 2016 meeting is available here.

Please contact Chris Derksen (chris.derksen[at]canada.ca) and Gerhard Krinner (gerhard.krinner[at]ujf-grenoble.fr) for more information.

In memoriam: Andrew Slater

Our very active steering group member Andrew (Drew) Slater passed away in the beginning of September 2016. Drew was 44. Like all his numerous colleagues, we will miss his great scientific contributions, and like everybody who knew him, we will miss his continuous good mood. A memorial website for Drew has been set up at http://www.forevermissed.com/andrew-slater/#about.

end of accordion