Community Snow Observations - Help us out!

Backcountry Enthusiast,

We are writing to share with you information about the Community Snow Observations (CSO) Project and to invite you to participate. The CSO project is an initiative funded by NASA to improve understanding of snow variability in complex terrain.

The CSO team uses snow depth measurements collected by Citizen Scientists traversing backcountry environments. These data are incorporated into numerical model simulations of snowpack and runoff processes. By sampling in a wide variety of locations and elevations, these data provide valuable spatial coverage that traditional fixed stations (SnoTel) are unable to provide.

The data are easy to collect, requiring only an avalanche probe, a smart phone equipped with the Mountain Hub app (available for IOS and Android devices) and enthusiasm from backcountry users like you. Tutorials on how to participate are provided at the CSO website and our Twitter feed will keep you up to date on the latest developments and modeling results.

Thank you for your time and, as the snow season kicks off, we hope you'll take a moment to contribute to our collective understanding of the snow environment.

David Hill, on behalf of the CSO Team.

How do I measure snowpack deeper than my probe?

Great question. Short answer; you don't. Or, you get a longer probe. Our modeling will greatly benefit from measurements in areas where the snow depth is less than your maximum probe length.

Thank you,
David Hill, on behalf of the CSO team.

right!

Hey People!

If you're interested in seeing how these measurements help, please visit the link below.

The first figure plots modeled snow-water-equivalent (SWE) vs. observed SWE at a SnoTel site in our model domain. Points are colored by 'day' in the water year. This is for a 'best calibrated' run. If the model was perfect, all points would be on the 45 deg sloping line. As you can see, the model builds way too much snow. This can be due to many things, in this case it is most likely a result of our choice of weather reanalysis product (MERRA). Basically, the weather product is too cold and wet.

The other figure shows the model when it is run with assimilated data from citizen scientists. Note how the points are much much closer to the 45 deg line. The root-mean-square error in modeled SWE is cut from 22 cm to 3 cm.

Science stoke!

www.flickr.com/photos/128442461@N06/albums/72157690574660205

Great question. TL:DR --> measure with probe vertical.

If you want a more detailed response, here it is. We are ultimately interested in accurate models of snow (or water equivalent) volumes over large areas. We discretize the real world with a regular horizontal grid. In each grid cell, we then have a depth. So, we can quickly numerically integrate over our model domain to get volume.

Let's use the example of a flat football field that is 100 m by 50 m (metric, I know...). If there is a uniform 1 m layer of snow, then we see that we have 5000 m3 of snow. Now, let's lift up one of the end zones and tilt the field to a 45 deg angle. Because we all ski gnarly steeps like that. Now, the effective horizontal dimensions of the field are 71 m by 50 m. So, the 'horizontal projection' of the area is a bit over 3500 m2.

So, if you used a probe and measured perpendicular to the snow surface, you'd see a depth of 1 m. When we multiple that by the horizontal projection of the area, we get 3500 m3 of snow, not the 5000 that is present on the slope. Remember, this is all due to our model 'grids' being horizontal planar surfaces.

Instead, if you used your probe to make a vertical measurement, you'd find a depth of 1.4 m. Multiply that by the 3500 m2 area and you get the correct answer.

So, vertical probes, please!

Regards,
Dave Hill.