❄️ SNOTEL Leaders

Variables

• Snow Depth: The actual height of the snowpack from the ground up, measured by an ultrasonic sensor.
• Snow Water Equivalent (SWE): The weight of the snowpack expressed as a depth of water. SWE represents how much water is stored in the snow and is a critical variable for water resources monitoring. Measured using a snow pillow.
• Precipitation: Cumulative water-year precipitation.

Concepts

• Water Year: A redefinition of the year as lasting from October 1 to September 30. Used by scientists and water managers, water years align better with the hydrologic cycle of snow accumulation starting in the fall and snow melt in the spring. If you perform analysis with calendar years, you end up looking at data from two different winters.
• Peak Snow Depth: The maximal snow depth observed for a given time period. In our case, the consistency metric looks at peak snow depths per water year.
• Z-score: A statistical measure of how many standard deviations above or below the mean a data point is. Z-scores tell you how anomalous a piece of data is.
• SNOTEL Overview from OpenSnow
• SNOTEL Overview from NIDIS

Data Validation

Data is passed through automated checks including...

• Range Filters: Removes physically impossible or unlikely values like negative depths or 10 meter SWEs.
• Cross-variable Checks: Cross-references values like Snow Depth and SWE or precipitation accumulation and SWE to flag unexpected relationships.
• Day-over-day Spikes: Flags sudden and improbable single-day shifts in snowpack for further review.
• The system supports filters (completely remove data points) and flags (display a warning but keep the data). The system is designed for extensibility and only requires a Polars expression + a few pieces of metadata to define a new check.
• Validation check definitions