Hydrology, Cryosphere & Earth Surface Project Update

Crowdsourcing Snow Depth Data with Citizen Scientists

A new project harnesses the power of the winter backcountry recreation community to gather data that are vital to understanding snow, from winter hazards to water resources.

By , Gabriel J. Wolken, Katreen Wikstrom Jones, Ryan Crumley, and Anthony Arendt

For many people, the arrival of winter focuses their attention on snow—its depth, distribution, and surface characteristics—so that they can plan for their next ski, snowshoe, or snowmobile adventure.

For many environmental scientists, winter’s arrival kicks off a different kind of snow season: one spent tracking snow cover and snow’s impact on ecosystem dynamics, hydrological systems, and glacier health. Knowing how much snow is on the ground at any given time allows scientists to predict trends in water supply for agriculture and hydropower and provides crucial information about where and when we might encounter snow-related hazards such as avalanches.

But what if those who play in the snow could help those who study the snow? A new NASA-funded project called Community Snow Observations (CSO) hopes to facilitate exactly that. It aims to blend the activities of both scientists and recreationalists to broaden and improve our understanding of snow.

Measuring the Snowpack: Challenges and Limitations

Snow processes are highly variable in space and time, making it difficult to make accurate estimates of total snowpack amount and distribution. In the western United States, a combination of automated snow telemetry (SNOTEL) stations and snow courses are operated by the Natural Resources Conservation Service. These are excellent sources of data, but they are limited in number due to funding constraints, and they tend to undersample complex topography and the highest elevations because of access constraints.

Airborne and satellite assets are able to remotely sense the snowpack and overcome some of the spatial sampling issues of in situ measurements. For example, NASA’s Airborne Snow Observatory uses lidar to measure snow depth, and the Moderate Resolution Imaging Spectroradiometer (MODIS) sensor on NASA’s Terra satellite is able to measure snow-covered area.

However, these observations may still lack the temporal and spatial resolution and coverage needed to accurately account for all patterns of seasonal snow behavior, such as snow depth changes after a large storm.

Can the Public Help?

To address these challenges, the CSO project began in February 2017 when researchers began to wonder if the public could play a valuable role in collecting unique data on snow depth.

The CSO team specifically identified the winter backcountry recreation community—skiers, snowboarders, snowshoers, snowmobilers, and avalanche professionals—as ideal candidates to contribute since those individuals often travel (1) over long distances, (2) up to very high elevations, and (3) far away from roads and other infrastructure such as ski lifts and gondolas. All of these characteristics describe the exact areas that tend to be poorly represented by fixed observations and are therefore highly desirable to measure.

Citizen Science: History, Challenges, and Opportunities

The idea of crowdsourcing data from citizen scientists is far from new. A look at prior citizen science projects shows that they bring a unique set of opportunities and challenges.

One key advantage is that the measurements have zero cost: They come from volunteers who possess the resources and skills to travel to the field, collect the data, and transmit them to the project team. In addition, there is some evidence that citizen science democratizes access to science and encourages scientific literacy in the greater public.

Regarding challenges, by its very nature, citizen science is decentralized and unstructured. Measurements are opportunistic and depend upon decisions (routes taken, days traveled, etc.) made by the citizen scientists themselves. The project team can offer suggestions and guidance but, ultimately, must cede control of the experimental plan to the citizen scientists themselves.

Data quality control presents another challenge. Protocols can be developed and tutorials provided, but in the end, the project team needs to accept that measurements are coming from a diverse body of contributors with differing levels of experience with data collection.

How CSO Works

A researcher measures snow-water equivalent in Alaska
David Hill uses a snow coring device to measure snow-water equivalent (SWE) near Thompson Pass, Alaska. Since measuring SWE takes specialized equipment like this, citizen scientists measure snow depth for the project instead. Credit: Ryan Crumley

To maximize the success of the CSO project, the team adopted a few criteria. First, measurements should require minimal equipment. Second, measurements should be fast and easy to make. Third, the flow of data from citizen scientist to project team should be highly automated.

To meet the first two criteria, the CSO team decided to focus solely on snow depth measurements, rather than snow-water equivalent (SWE) measurements. Measuring SWE takes specialized equipment and a moderate amount of time, whereas measuring snow depth takes mere minutes and requires only a device with graduated markings.

Although in principle a ruler or tape measure could be used, many backcountry users already carry a perfect measuring stick: a collapsible avalanche probe. Along with an avalanche beacon and a snow shovel, these probes are part of an essential safety kit. Roughly 2–3 meters in length with ruled markings at 5-centimeter intervals, the common avalanche probe is ideal for measuring snow depth.

The team developed online tutorials to guide contributors through various best practices to follow since there are some nuances to making a good snow depth measurement (e.g., avoiding disturbed areas). With this bare minimum of training and practice, however, it takes only a minute to assemble a probe, make a measurement, restow a probe, and be on your way.

To meet the third criterion, the CSO team partnered with Mountain Hub, which developed an outdoor-oriented, community-fueled smartphone application in 2015 that now has a digital audience of over 100,000. Working together, CSO and Mountain Hub added a “snow depth” field to the portion of their app that allows users to report on snow conditions. The app now logs location, time, and snow depth and automatically transmits these data to the CSO science team. These data are also shown online in real time.

Collaborating on an app that already exists and is widely used by the target audience, rather than developing a new one, has been advantageous in reducing development cost, project spin-up time, and the “app clutter” that overloads many phones. More recently, CSO has also partnered with a software program called SnowPilot to obtain depth data directly from avalanche professionals.

What Happens to the Data?

The snow depth data (available for interactive viewing here and for download through an application programming interface here) are being used by the CSO team in several ways. First, the data are used operationally to validate snow distribution products derived from airborne snow remote sensing data (i.e., lidar and photogrammetry).

Second, the data are assimilated into snow hydrology models to help scientists better constrain the amount and distribution of SWE throughout the snow season. Every time a measurement is made, the error between the model’s predicted depth and the actual depth at the measurement location is used to steer the model back toward the ground truth.

A scientist measures snow depth with an avalanche probe
Katreen Wikstrom Jones uses an avalanche probe to measure snow depth near Turnagain Pass, Alaska. Credit: Gabriel Wolken

The citizen scientists therefore play a valuable role in providing better estimates of snowpack storage and, ultimately, spring runoff. Initial tests of this assimilation strategy have shown that the contributed snow depth measurements dramatically reduce model errors.

Third, increasing the accuracy of snow hydrology models can help improve satellite estimates of water storage. The Gravity Recovery and Climate Experiment (GRACE) and GRACE Follow-On (GRACE-FO) missions use satellites to measure gravity anomalies. By incorporating various models of physical processes (tides, isostatic adjustment, etc.) it is possible to isolate a certain component of overall mass (e.g., glaciers) from the GRACE data. The CSO team will contribute to these efforts by supplying their high-resolution hydrological model results.

Looking Ahead

The first year of the CSO project was extremely successful. Beginning near Valdez, Alaska, with several intensive measurement campaigns, the project spread rapidly through the western United States, with several thousand measurements obtained.

A map shows snow measurements in the western United States
A map of the locations of CSO measurements across the western continental United States from October 2017 to April 2018. Credit: David Hill

In April 2018, NASA announced that CSO was selected as one of six projects to receive continuing funding for the next 3 years from its Citizen Science for Earth Systems Program. Our team is busy analyzing the data and developing strategies to scale up CSO efforts moving into the 2018–2019 winter season.

There is no crowdsourcing without the crowd, and CSO hopes to see more and more citizen scientists out in the snow this season. Project participants have remarked that making repeated measurements for CSO has helped them learn about how snow is distributed and how it evolves over the season.

Snow is a winter playground for many but an important water resource for all. Lend CSO a little bit of your time, and turn your winter experience into data for science.

—David Hill ([email protected]), Oregon State University, Corvallis; Gabriel J. Wolken, University of Alaska Fairbanks; Katreen Wikstrom Jones, Alaska Department of Geological and Geophysical Surveys, Fairbanks; Ryan Crumley, Oregon State University, Corvallis; also at University of Washington, Seattle; and Anthony Arendt, University of Washington, Seattle

Citation: Hill, D., G. J. Wolken, K. W. Jones, R. Crumley, and A. Arendt (2018), Crowdsourcing snow depth data with citizen scientists, Eos, 99, https://doi.org/10.1029/2018EO108991. Published on 03 December 2018.
Text © 2018. The authors. CC BY 3.0
Except where otherwise noted, images are subject to copyright. Any reuse without express permission from the copyright owner is prohibited.