OSD/GCW-CNT1/Doc. 3.1.3, Page 1/13

OSD/GCW-CNT1/Doc. 3.1.3, Page 1/13

WORLD METEOROLOGICAL ORGANIZATION

______
GLOBAL CRYOSPHERE WATCH (GCW)
CryoNet Team Meeting
First Session
Reykjavik, Iceland
20-22 January 2014 / OSD/GCW-CNT1 / Doc 3.1.3
(16.01.2014)
______
ITEM: 3.1
Original: ENGLISH

GCW SNOW WATCH WORKSHOP

summary

(Submitted by Barry Goodison on behalf of GCW Snow Watch leads, Kari Luojus and Ross Brown)

ISSUES TO BE DISCUSSED:
The document provides a summary of the recommended actions from the GCW Snow Watch workshop and the current status of these actions.
DECISIONS/ACTIONS REQUIRED:
1. Identify mutual interests and actions between CryoNet and Snow Watch requiring further action.
2. Recommend issues for which CryoNet could/should take the lead for implementing, and those where CryoNet would partner with Snow Watch on implementation.
3. Identify interested experts to contribute to all Snow Watch recommendations.

OSD/GCW-CNT1/Doc. 3.1.3, Page 1/13

GCW SNOW WATCH WORKSHOP

1. Environment Canada hosted a first workshop on implementing a Snow Watch component of GCW at Downsview, Canada from January 28-30, 2013. Twenty-eight scientists from nine countries (Austria, Canada, Finland, Italy, France, Norway, Switzerland, United Kingdom and USA) participated in this initial workshop to determine the current state of global snow monitoring, to identify critical issues affecting the ability to provide authoritative information on the current state of snow cover, and to initiate GCW Snow Watch projects to address priority areas. The workshop presentations, list of participants, participant questionnaire responses and rapporteur summaries are provided online through the GCW website ( or at WMO The final recommendations for priority areas and potential GCW projects are provided below based on the questionnaire responses and workshop discussions:

2.1. Final Workshop Recommendations

2.1.1 Improve real time flow and access to in-situ snow measurements:

This was rated as one of the highest priority items for GCW in the questionnaire results. Specific issues raised with respect to in-situ data included the lack of zero-depth reporting on the GTS network, non-reporting of snow depth observations by some countries, small numbers of US snow depth reports in SYNO messages, a need for standardized QC of snow depths from autostations, and the development of a global historical in situ snow data archive to support climate monitoring, model evaluation and process studies.

Action: Eric Brun (Meteo-France) agreed to lead a team to improve data exchange on the GTS, to demonstrate the advantages to NWP of zero-snow depth reporting, and to brief EC-PORS and get their support. Jay Lawrimore (NCDC), Sean Helfrich (NCDC) and Patricia De-Rosnay (ECMWF) agreed to contribute to this effort. In addition, the workshop identified the need for a data rescue project to identify paper and digital snow data that were not readily available in an existing archive. Barry Goodison, Julie Friddell (U Waterloo), Jay Lawrimore and Kari Luojus offered to help.

Update: Eric Brun reported that the team turned the snow workshop discussions into a document describing various aspects of the initiative. The most recent draft report (V4) is provided in Appendix A.It was forwarded to the Secretariat in September in order to get a formal endorsement as well as WMO contacts who should help with the technical issues of snow data dissemination. No response has yet been received even after a second mail a few weeks later.

This issue is relevant to CryoNet as delivery of data from the sites to users via WIS is an important issue. CBS and CCl will have direct interest in these two aspects.

2.1.2 Initiate a satellite snow products evaluation/intercomparison activity

There was widespread agreement amongst workshop participants that a comprehensive evaluation/intercomparison of snow cover products was a high priority for GCW given the increase in the number of products in recent years and the need to provide uncertainty estimates for assimilation of data into operational snow analysis schemes. For example, Drew Slater provided results of an intercomparison of gridded SWE products that revealed serious problems with SWE estimates obtained solely from SSM/I and AMSR-E brightness temperatures (SWE values unrealistically high over Siberia). These data have been used to evaluate climate models with erroneous conclusions published in the literature.

Action: Bojan Bojkov (ESA) agreed to lead a GCW intercomparison project of SE and SWE algorithms and products that followed well defined, and community accepted, protocols both for the generation of datasets and for the validation/intercomparison. He proposed an algorithm round-robin approach following the ESA Climate Change Initiative (CCI) using independent in situ validation datasets of quantified quality. The results would be formally reviewed by independent reviewers. Drew Slater (NSIDC), Chris Derksen (EC), Ross Brown (EC), Richard Fernandes (CCRS), Kari Luojus (FMI), Thomas Nagler (Enveo IT GmbH) and Dorothy Hall (NASA) agreed to participate in this project.

Update: ESA has offered to organize and support a “Satellite Snow Products intercomparison and evaluation EXercise – SnowPEX” project to be carried out by a team of international experts beginning in 2014. The project will intercompare and validate current global/hemispheric satellite snow products for assessing their quality and for better quantifying the uncertainty of long term trends of the seasonal snow pack deduced from satellite data. Thomas Nagler has led the preparation of a proposal for ESA.

Action: Richard Fernandes (CCRS) proposed a parallel initiative for a PI self-assessment of snow cover products to provide quick feedback on the strengths and weakness of current snow cover products to the user community. This may help avoid situations such as that described above where products with known weaknesses contribute to misleading conclusions. Kari Luojus (FMI) and Dorothy Hall (NASA) agreed to participate in developing this project.

Update: no update available

These above actions are relevant for CryoNet as independent and integrated CryoNet sites could serve as important independent data sources for such intercomparisons of not only snow but also other cryosphere components.

2.1.3 Development of hemispheric "snow anomaly trackers" for SE and SWE

Kari Luojus (FMI) presented a prototype anomaly tracker for near real-time monitoring of SWE and SE anomalies at global to continental scales following the model of sea ice anomaly tracking at NSIDC. These could be highly visible GCW tools for identifying and monitoring the development of snow cover anomalies e.g. below-normal winter snow accumulations, advanced snow melt. This information can be used to inform decision makers and the public, and also to promote media interest in snow cover. In order to be “authoritative” tracker products will need to be based on multiple products and be supported by a credible team of experts.

Action: Kari Luojus (FMI) agreed to lead a project to develop this concept with Ross Brown (EC), Chris Derksen (EC), Dave Robinson (Rutgers U) and Stefan Wunderle (U. Bern).

Update: The Snow Watch Group stimulated the development of new daily “trackers” for Northern Hemisphere snow extent and snow water equivalent. Snow trackers have been developed for GCW by the Finnish Meteorological Institute and Environment Canada. The FMI SWE tracker used SWE algorithms developed by GlobSnow, was initially developed shortly after the meeting and showed the NH snow mass to be well beyond the +1SD of the mean snow mass average for 1982-2002. These products are available in near real-time on the GCW website. Satellite, in-situ and operational NWP analyses contribute to the development of these snow products.

These products are relevant to CryoNet, since in-situ snow depth data are used in the algorithm to generate SWE. Hence standards, guidelines and exchange of data in real-time for snow depth measurements are very important.

2.1.4 Develop inventory of existing snow datasets

The development of a baseline inventory of snow-related datasets was considered an important priority for GCW. This would contribute to several of the activities mentioned above.

Action: Ross Brown (EC) and Julie Friddell (U. Waterloo) agreed to take the lead on pulling this together.

Update: Ross Brown reported that a draft inventory was being prepared for discussion at a GCW-related meeting in conjunction with this year’s Eastern Snow Conference (ESC).

2.1.5 Initiate activity to standardize snow-related nomenclature, and promote standards and best practices as a contribution to CryoNet

The standardization of observing practices and snow-related nomenclature was considered a high priority activity in the questionnaire particularly for snowfall and solid precipitation where practices vary widely between countries. This activity will contribute to the GCW CryoNet activity which aims to establish a network of cold region reference sites with sustained, standardized observing programmes for climate and cryospheric variables.

Action: Barry Goodison agreed to lead an activity with Thomas Nagler (Enveo IT GmbH ). Jeff Key (NOAA) volunteered to compile a list of snow measurement-related manuals and reference material.

Update: Jeff Key has led the compilation of standards and best practices for cryospheric measurements for GCW, including snow measurements. GCW is drawing on existing measurement methods where possible and where a scientific consensus has been or can be reached. An initial inventory of existing documents describing measurement practices is given on the GCW website, under CryoNet ( A glossary has also being initiated for GCW by Jeff Key and currently has over 2000 entries from several sources (see Both are in continuing development.

This action was seen as a contribution to CryoNet development and will also contribute to the work of the Infrastructure and Practices Team. Of course, CryoNet is concerned with all components of the cryosphere.

2.1.6 Follow-up meetings

It was agreed that the GCW Snow-Watch community needed to meet regularly and that it also needed to respond to snow monitoring issues in the Southern Hemisphere and Asia e.g. through regional meetings and regional initiatives.

Action: Stefan Wunderle (U. Bern) invited the GCW Snow-Watch workshop participants to meet in Bern in February 2014 as part of the 7th EARSeL Snow and Ice Workshop.

Action: Sean Helfrich (NOAA) indicated he was interested in hosting a small, focussed North American snow cover monitoring working group meeting in the fall of 2013.

Update:February 2014 was deemed to be too early for the next meeting, and a fall time frame is seen to be more suitable. A NA GCW-related meeting (1-day) will be held before or after the next ESC in North Carolina (early June).

OSD/GCW-CNT1/Doc. 3.1.3, Page 1/13

APPENDIX A

A Global Cryosphere Watch Initiative

- for improving in-situ snow observations and their access

- for rescuing/collecting historical in-situ snow data

Eric Brun / Meteo-France CNRM, Toulouse, France

Jay Lawrimore / NCDC/GCAD, Asheville, USA

Patricia de Rosnay / ECMWF, Reading, UK

Julie Friddell / University of Waterloo, Waterloo, Canada

Version 4, 11 October 2013

Rationale:

-Snow cover is a major component of hydrological resources in North America, Northern Eurasia, Mediterranean countries, and Asian and South-American river basins including high mountain regions.

-Snow cover is a major source of positive climate feedbacks (e.g., ice-albedo feedback).

-Snow cover is a very sensitive indicator of climate change.

-Snow cover strongly impacts surface meteorological conditions.

-There remain large uncertainties in snow water equivalent (SWE) and snow depth (SD) satellites retrievals. These uncertainties are significantly reduced when ground-based snow depth observations are assimilated.

-SD observations from synoptic weather stations are critical for the analysis of snow cover in numerical weather prediction (NWP) and reanalysis systems.

-Synoptic stations make up only a portion of available SD observations which also are provided by many climate observing networks.

-SD and snow density are critical variables for the seasonal and long-term evolution of frozen ground and permafrost.

Consequently, present, past, and future ground-based SD and SWE observations (as well as snowfall, SF) are of utmost importance for operational water resources monitoring, for climate change detection, for satellite snow products evaluation, for snow model evaluation, for meteorological reanalysis evaluation, and for climate model evaluation.

1. Improvement of the operational ground-based SD and SWE observations

During the GCW SnowWatch Workshop held in Toronto, January 2013, several issues were identified:

• Issue 1: In snow covered regions, many SYNOP messages omit SD observations when snow is not present on the ground (see Figure 1 showing the seasonal variation in the number of SD observations extracted from the ECMWF MARS server). It is therefore very difficult to unambiguously identify the onset and vanishing dates of the snowpack, making it difficult to analyse the snow cover in NWP and reanalyses systems, with an asymmetry between snow and snow-free periods. In addition, scarcity of SYNOP reports of SD (even in snow covered conditions) in large areas such as the US (Figure 2), UK, China, and southern hemisphere is an issue for NWP and other near-real time applications.

• Issue 2: Occasional errors occur in SYNOPSD observations, which impact the analysis of SD in NWP systems (see Figures 3 and 4) and in some satellite-based snow products.

• Issue 3: SD observations from automatic stations often show an offset of a few centimetres which limits the possibility to detect the onset or the vanishing dates of the snowpack.

• Issue 4: Scarcity in SWE observations collocated with SD observations, which are critical for creation of many products (hydrology / NWP analysis and others) as well as for the evaluation of satellite-based products.

To analyse and possibly solve these issues, the GCW task team has formed a working group with the following objectives:

• Issue 1: An analysis of the current situation, in links with WMO Commission for Basics Systems (WMO/CBS) and some stakeholders (climate data centres, ECMWF, National Meteorological and Hydrological Services (NMHS), …), should lead to a rapid identification of the reasons for this issue. WG1 should identify the solutions to be implemented, probably based on the practices already existing in some countries, such as the BUFR template described below. Large regions and countries show extremely sparse SYNOP stations reporting SD. For some of these areas, however, national networks provide NRT data that could be made available on the WMO Global Telecommunication System (GTS). In Europe, an initiative from ECMWF consisted of developing a BUFR template (see Annex 1) that ECMWF member states are encouraged to use to report their national network SD data at 06UTC. European national network observations have been available on the GTS since March 2011. To this date six countries put their national SD data on the GTS in addition to SYNOP data (Sweden, the Netherlands, Denmark, Romania, Hungary, and Norway) and the data are used in the operational ECMWF snow depth analysis. This BUFR template is available to be used by WMO Member states and SD data providers to report and put on the GTS network SD data for the NWP and NRT applications communities.

• Issue 2: In priority and starting from recent examples, an analysis of the real-time availability of SD data should lead to proposals for making data from synoptic and climate networks more widely available over the GTS; e.g., the US COOP Network and others globally. This should be addressed in close relationship with WMO/CBS. With a lower priority, WG1 should make proposals for the implementation of quality control (QC) algorithms. QC algorithms could be implemented either at the level of the observation producers or at the level of data centres or users. These algorithms could be derived from those to be developed for the action on SD data rescues (see Section 2 below).

• Issue 3: Building on the experience of some operators of automatic snow observations networks, WG1 should analyse the current practices and promote a wider use of QC and correction algorithms.

• Issue 4: WG1 should analyse the current situation, in links with CBS and the concerned NMHS, which should lead to the promotion of regular ground-based SWE observations, at least over a reference network.

Tentative membership of WG:

Co-chairs: Eric Brun (France) and Jay Lawrimore (USA)

Julie Friddell (Canada)

Patrica de Rosnay (ECMWF)

Ioannis Mallas (ECMWF)

Rick Fleetwood (Canada)

A representative of WMO/CBS (tbd)

An expert in observation networks from Russia (tbd)

An expert in observation networks from China (tbd)

An expert in climate data processing (Germany) (tbd)

.....

2. Rescue and/or collection of ground-based SD and SWE historical data

Some ground-based SD and SWE quality-controlled data sets exist and are extensively used by the scientific community (North-America, former Soviet Union / Russia, …) but there is an immense gap between user needs and current availability of snow data of all types, due primarily to the following reasons:

-It is very difficult to access historical SD data from many European and Asian countries.

-There are large gaps in historical records of SD and SF throughout large parts of the world.

-SD quality-control is relatively weak or absent.

-Relevant meta data are unavailable.

It is likely that a huge amount of past SD observations (such as those from SYNOP messages which may be accessible through the US National Oceanic and Atmospheric Administration/National Climatic Data Center (NOAA/NCDC) as well as SD observations from other climate networks) could be rescued from paper archives or collected from existing digital datasets held in NMHS’s. These observations could be subsequently quality controlled and made accessible through the GCW portal.

This represents an important effort which could be led by GCW relevant teams in partnership with the concerned NMHS and data centers (including ECMWF). Different stages are necessary: