From The Montrose Press (Michael Cox):
The primary tool currently in use to measure snowpack in the Western United States is SNOTEL. We all rely on the SNOTEL website to see what’s happening during winter in the Rockies. But, you may be surprised to learn that the SNOTEL (SNOw TELemetry) has been missing the mark in its automated reading of snow depth in the Western US. How do we know that? Because, there is a new tool – actually an old one, repurposed – that could enhance greatly the accuracy of the 732 SNOTEL stations currently being used for the critical purpose of measuring snowpack in the mountains to help water managers forecast the potential runoff.
The solo SNOTEL system was as good as it got for 50 years when it came to measuring snow in the mountains. The system of sensors that measure snow depth and the amount of water contained in the snow was put into use back in the 1970s. It has not been updated since then, although some stations were added in the 1980s. SNOTEL measures two primary parameters, snow depth and density. Density tells us how much water is in the snow. It does this by sensing the weight of the snow on something called a snow pillow. The pillow is about eight feet square and as the snow builds up, it gets weighed. That number and the depth at the station are reported to the system as what we call the snowpack.
SNOTEL actually functions pretty well up to a point. The biggest drawback with it is the minuscule sampling of a vast area of snow production. The 732 stations are spread out through the mountain snow regions of all the Western states, including Alaska. That area is 1.76 million square miles, of which about a third is mountainous and has snow pack. That means there is a SNOTEL station for every 800 square miles of mountain terrain. Some of the stations are not as accurate as they need to be because of location. Some terrain, where extraordinary snow accumulation occurs, such as the bottom of an avalanche chute, never get measured because they are below the altitude level where SNOTEL stations are located. The avalanche-prone San Juans may have much more snow than we ever knew.
Given the increasingly critical nature of determining even short term snow inventories, people like John Lhotak, an operations hydrologist with the Colorado River Basin Forecast Center, told a press meeting, “SNOTEL is the best network we have, but there are definitely shortcomings.”
Enter LIDAR. LIDAR is one of those pseudo-acronym things that the lab guys and bureaucrats love. This one stands for Light Detection and Ranging.
Quite simply, if you flew over the mountains without snow on them and determined the height (compared to sea level), and then flew over and scanned them when the snow is in place, you would simply deduct the original snow-less height from the snow packed image and “voila!!!” you get the snow depth of the whole mountain almost to within centimeters.
Sounds simple enough, but the data crunching is mind numbing. All the data points from the ground-only image must be overlaid with the image taken with snow on the ground. The measurement points are chosen and then comes all the subtraction and interpolation. The people like Jeffrey Deems at the National Snow and Ice Center and Sam Tyler at Utah State University (and their teams) have developed the computer tools to breakdown the gigabytes of data collected to simple usable terms.
The whole concept was first tested in California’s Sierra Nevada Mountains eight years ago. The dry model of the mountains was made by flying at 20,000 feet in a straight back-and-forth pattern. After some storms passed the location, the team went back and flew the same pattern at the same altitude. The resulting 3D images were a precise measurement of the snow on the ground. Tyler’s team also did a test of the system near Logan, Utah, at about 8,000 feet…
The Airborne Snow Observatory (ASO) folks tell us, “We see it as moving from a sparse-point base network (with SNOTEL) to a system that can map the entire snow pack in a river basin,” Jeffrey Deems said, “It is really an enabling technology.”
In 2013 the ASO tested the system on selected sections of the Front Range, Gunnison Basin, Rio Grande Basin, and Uncompahgre watershed. Deems said, regarding the SNOTEL numbers, “We were missing a lot of the picture. We need to fix that.”
What the tests revealed was that in the Rio Grande Basin, for example, the forecasts were way off, reporting as much as 50% less snow and water than what was actually on the ground. That makes accurate forecasts and water use management for that basin impossible…
But the bean counters aren’t so sure. First of all, flying several thousand miles back and forth over the Colorado peaks costs a lot of money. The tab for flying for the new imagery on a regular basis could cost $400,000 a year or more, according to Frank Kugel, director of the Southwest Water Conservation District. Is the return on investment really there?
Also, everyone in the water biz seems to agree that we will still need SNOTEL. It is currently the only tool for proofing the accuracy of the LIDAR images and vice versa. It is also the best tool for the density issue. For the time being, people like Deems think using SNOTEL in tandem with LIDAR is the right way to get the best measurements. Rather than replacing SNOTEL, Deems would opt for even more SNOTEL stations…
Deems said [February 6, 2020] that the cost of LIDAR seems justified when you consider the cost of a bad forecast. It is no secret that the low estimate on the Rio Grande in 2013 translated into millions of dollars of water misused after the forecast. Making the investment available for better measurements seems like a no brainer…
Meanwhile, the Colorado Water Conservation Board has already decided to invest $250K in 2021 for flights to measure the Gunnison Basin, of which the Uncompahgre River is a part.