Snow and Precipitation FAQ
Snow and Precipitation Frequently Asked Questions
The estimated data in historic monthly SNOTEL SWE tables (Here’s an example from Atlanta Summit SNOTEL) are back generated values for the SNOTEL site based on measurements from a highly correlated snow course with a longer record. In many cases the SNOTEL was built very close to an older manually measured snow course. For the first 10 years or so after the SNOTEL was installed the snow course was still measured until a reliable regression relationship between the two locations could be developed. If a highly correlated relationship was found (R^2 > 0.90), then that relationship was used to back estimate SWE data for the SNOTEL site based on the historic snow course measurements. At that point the snow course was generally discontinued.
Cumulative precipitation should only increase through a water year, however we allow small decreases in raw precipitation data (less than -0.5 inches) so that it passes through our automatic quality checks. This allows the data to be included in our data reports. For large decreases (greater than -0.5) the data is flagged suspect and will show up as a -99.9, those values are kept out of data reports. The reasons you might observe decreases include...
1. The voltages that our electronic pressure transducers use to measure precipitation and snow water can flutter up and down with temperature changes. This is particularly evident in hourly data, since diurnal temperature changes cause greater fluctuations than readings taken on consecutive midnights, when we measure daily data.
2. There might be an air bubble in the plumbing line leading to the transducer. Temperature changes cause the air bubble to expand and contract and this is another reason for measurement flutter. Bleeding the air out of all plumbing lines is one of our common field maintenance practices.
3. The data you are looking at hasn't been manually quality checked yet. We do weekly edits each Monday to smooth these kinds of flutter out of the daily data. Generally in a day or two the data will rebound to where it was and we'll edit out the low value the following Monday. The hourly data is raw and it does not undergo this weekly quality checking process so decreases remain in those data sets. If a precipitation decrease is observed in daily data that is more one water year old, it should be brought to the attention of the snow survey as it should be corrected.
4. The final case is there is something wrong at the site. There are various problems that can cause a precipitation gage to decrease. These include leaks caused by an animal such as a bear or mouse chewing a hole in the plastic plumbing, or a human vandalizing the gage with bullet holes. When a leak is suspected we schedule a repair trip to the SNOTEL site to fix it. Until the repair is made we edit the data using nearby sites or during the winter we'll estimate precipitation using the snow pillow.
Unfortunately there is not an easy or quick way to query the database for the summer snow-free date each year at SNOTEL sites. The best way to do this involves a labor intensive process of scrolling through the annual historical SWE (snow water equivalent) daily 'tables' and just jotting down the day for the years that interest you.
Navigate to a SNOTEL site page from the clickable map, for example Bunchgrass Meadows SNOTEL. To find these tables, click the "Daily" link under "Historical" on the "Snow Water Equivalent" row. These tables are the most compact format of the daily data, with 31 lines per year. The linear "tab formatted" version, with 365 lines per year, are best for spreadsheet applications.
Some sites accumulate more snow water than precipitation year after year. This has to do with the specific characteristics of the site. Schweitzer Basin is one of a number of sites where the snow pillows are in a location where they tend to accumulate more snow than what falls in the rain gage. This is due to the fact that wind can cause snow drifts on the pillow and because the precipitation gages tend to under catch precipitation when it's windy. Other sites where this phenomena is common include... Bogus Basin (Boise basin), Two Ocean (Upper Snake basin, WY), Blind Bull (Upper Snake basin, WY), and Grand Targhee (Teton basin, WY). Most of these sites are at high elevation, exposed locations where it's windy and the pillows are a sheltered just enough that the snow tends to drift on to the pillows. Although it might seem like we should have chosen a better location, in many cases that wasn't possible because the ideal opening didn't exist at the elevation where we needed data.
On the flip side we also have a few sites where the precipitation tends to gain on the snow water. Classic examples of this include: Prairie (Boise basin) and Humboldt Gulch (Coeur D'Alene basin). Sites that do this generally are at lower elevations where deep snowpacks are not common. These sites often see rain on snow events mid-winter, these cause snow to melt and runoff to occur. As might be expected due to the nature of the weather in Northern Idaho, this is more common at sites in that part of the state since overall elevations are not as high and temperatures tend to be more moderate.
It should be noted that the majority of our sites are in ideal locations where snow water gains are pretty much one-to-one with precipitation accumulation during most winters. Examples include: Galena (Big Wood basin), Jackson Peak (Boise basin) and Mosquito Ridge (Coeur D'Alene basin). However even these sites occasionally will see a storm or a winter where the winds approach from an unusual direction or the temperatures are warmer than normal and the one-to-one relationship doesn't occur.
The good news is that since each site behaves in a consistent pattern year after year, each SNOTEL site allows us to compare the current year versus the historical record in a reliable manner. This allows us to make accurate water supply forecasts and provide users with a meaningful percent of average snow water or year to data precipitation.
A snow index is a tool that allows you to compare the snow water amount for a specific basin with the historic period. Snow indexes are calculated by summing the snow water amount for the same list of SNOTEL sites for each year. Sums represent the first of month value. These sums can then be sorted (using a spreadsheet) to rank this year versus history to see how dry or wet the basin’s snowpack is this year.
Please contact someone on the snow survey staff.