MesoWest has been designed for use by National Weather Service meteorologists and other professionals for protection of life and property. Comments/suggestions for improvement may be sent to atmos-mesowest@lists.utah.edu.

MesoWest currently does have a few quality control methods that are applied to incoming observational data.

For all stations, quality control begins when the data are processed into MesoWest. These primarily consist of "range checks" which ensure any extraneous values are flagged as Suspect. Click here to view the maximum and minimum range check values associated with each variable. If the observation time appears suspect (e.g. a "future" observation where the time reported is well ahead of the current time), the observation will be flagged as Suspect Time.

The next two steps involve additional statistical checks for stations with the available observations. First, a regression statistical check is conducted on available temperature, dew point, and pressure data. The process for this check is explained in more detail here. Second, a different statistical check is then conducted on available wind observations using the previous 3 days of data. This check essentially looks for stagnant wind observations (e.g. a long period of observations with the exact same wind direction, speed, or gust) which likely indicate issues with the wind data.

For stations that are part of the SNOTEL network, quality control flags of 'valid' or 'suspect' are provided to MesoWest for each variable reported at a station. These are also utilized by MesoWest.

The majority of MesoWest products utilize the following designations and color scale to identify the current state of data quality. Additional information on the flags and keys can be found here.

Designation	Description
OK (black) OK (green)	Data has passed all Quality Control checks.
Caution (orange)	Some data has been flagged by the one of the statistical checks. This data should be used with caution and examined by the user.
Suspect (red)	Some data has not passed the range checks conducted on the variables. It is recommended that this data not be used.
Suspect Time (purple)	The reported time of the observation appears suspect, such as an observation where the reported time is ahead of the current time.

Please keep in mind that whenever quality control information is not available (often N/A will be listed where quality control information should be), all data should be used with caution. A good practice is to check the Station Interface for an individual station to view individual observations. Also, note that a Caution designation does not necessarily mean the data is bad, since there may be local conditions at a site which cause it to vary from its neighbors.

Additionally, the quality control flag listed is not variable-specific, so it is important to examine the actual station data to see which variable(s) may be causing a change in the station's quality control flag. Results of the statistical checks can also been seen on the Data Quality Plots. Typically, only one variable is responsible for a Caution or Suspect flag.

On some products, the station's name and/or elevation may be colored instead. These are used to indicate that the station location is Uncertain or Incorrect. Extra caution should be applied here, since the data may be valid but the station may really be in another state, for example.

*Important Notes:

1) Recall that the 5-Day Maximum and Minimum Summary also uses a quality control check for the amount of data, in addition to the quality control checks applied to the data itself. "Caution" or "Suspect" values may be due to a marginal or inadequate amount of observations, in addition to potentially erroneous values. See the Maximum and Minimum Summary Help Page for more information.
2) Recall that the Short-term Precipitation Monitor and Long-term Precipitation Summary products are limited to gross checks only for quality control. The best method to verify precipitation would be to examine the data for the station itself.