
Interim Issue Resolution: Air Temperature Data Bias
Air Temperature Bias Correction
Beginning in the 1990s, the YSI 44019A thermistor with the 44211A thermilinear package (the extended-range YSI temperature sensor) replaced the YSI 44230 temperature sensor as the default temperature sensor used at SNOTEL sites because of its ability to measure a broader temperature range. Air temperature is calculated from sensor output voltage using a linear least-squares regression algorithm created by the Snow Survey and Water Supply Forecasting Program*. Use of the linear least-squares regression algorithm results in a data bias.
* Air temperature data from Alaska automated sites also contains a similar but different bias. Use of a different equation is the source of Alaska’s temperature data bias. Correction of Alaska air temperature data will require the use of a separate debiasing equation. The Alaska debiasing equation is also under development.
Bias correction:
- Resolves bias introduced by inadequate sensor output conversion equation. The inadequate equation was erroneously used in place of the manufacturer suggested equation.
- Will not take into account or correct for environmental factors (e.g., sensor shield, shading, or radiative heating).
- Applies to all erroneous data across the SNOTEL network for which the sensor and equation were applied.
Metadata
The metadata in the file provides established dates on which known biased air temperature data were collected. The biased data are currently available in the SNOTEL database.
⇒ Download Air Temperature Bias Correction Metadata
Correction Equations
Note: Correction equations and guidance are in development.