Some Snow Spotter Results

Since launching Snow Spotter, we have gone through nearly 500,000 classifications on 50,000 images classified by over 8,000 volunteers! The results from citizen science classifications have been used in snow hydrology research to evaluate the canopy-snow interception component of our hydrologic models. Below are a few examples of sites on the left, with their corresponding data from the Snow Spotter classifications of snow in the canopy on the right, where snow in the canopy is in blue, no snow is in white, and no observations are in grey. This analysis is documented in our newly published peer-reviewed manuscript, Lumbrazo et al. 2022.

Evaluation of hydrologic models often focuses on snow on the ground, however the dataset created from these classifications are used to evaluate if our hydrologic model simulations are capturing the timing of snow in the canopy correctly. Below is an example of the interception dataset from Niwot Ridge, Colorado used to evaluate the hydrologic model, in black, showing snow water equivalent (SWE) in the canopy, mm.

The results from these analyses are reported in Lumbrazo et al. 2022. The first Snow Spotter dataset is published on Zenodo (Lumbrazo et al., 2022).

In this study, the snow spotter dataset helped us identify specific interception processes that our canopy-snow interception and unloading models are not correctly capturing, see the figure below.

The model simulations, shown in green, blue, and red show when our hydrologic models are modeling SWE in the canopy, which is compared to the interception observations from this study, in blue and white on the same plot at Mesa West in Colorado. Below that are the corresponding wind speed, temperature, and relative humidity for this interception event.

Rime Ice in the Canopy

This analysis highlights an example of rime forming in the canopy (images A—D). Rime forms when supercooled water droplets in clouds (image B) are carried by strong winds to engulf terrain (Whiteman & Garibotti, 2013). Rime formation often occurs with high winds, high humidities, and air temperatures within a specific range of -2 and -8 C (Whiteman & Garibotti, 2013). Photographs from this event at Mesa West show us that rime is forming in the canopy and remains in the canopy for many days after the initial interception event, while most of the model simulations unload snow form the canopy too quickly. Rime forming in the canopy happens a number of times throughout the season at Mesa West, which causes snow to stick to the canopy for a longer time (specifically during times with high wind speeds) when models would typically unload the canopy-snow.

The analysis from this first dataset from Snow Spotter sheds light on physical processes in hydrologic models that are often overlooked. The next steps are to determine if citizen scientists can characterize rime in the canopy! We are using Mesa West as a test site, so check out the riming Mesa West workflow to see if you can identify rime in the canopy for us and improve the modeling for the canopy-snow processes.

References
Lumbrazo, C., Bennett, A., Currier, W. R., Nijssen, B., & Lundquist, J. (2022). Evaluating multiple canopy-snow unloading parameterizations in SUMMA with time-lapse photography characterized by citizen scientists. Water Resources Research, 58, e2021WR030852. https://doi.org/10.1029/2021WR030852

Lumbrazo, C., Mozer, M., Currier, W. R., & Lundquist, J. (2022). Snow Spotter Canopy-Snow Interception Dataset #1(Version 1) [Dataset]. Zenodo. https://doi.org/10.5281/zenodo.5918637

Whiteman, C. D., & Garibotti, R. (2013). Rime mushrooms on mountains: Description, formation, and impacts on mountaineering. Bulletin of the American Meteorological Society, 94(9), 1319–1327. https://doi.org/10.1175/BAMS-D-12-00167.1