NovaStar Data Model / Calibration
Calibrations indicate how to convert raw point data values to scaled values with engineering units suitable for analysis and information product.
Overview
Each point can have one or more calibration records,
which provide constants to convert raw sensor values to scaled values with engineering units,
using the following equation.
For example, precipitation bucket tip counts are converted to inch or mm units
and pressure transducer voltage is converted to water level.
A calibration has an effective time,
which allows multiple calibrations over time to indicate changes in sensor characteristics
and the physical configuration.
ScaledValue = RawValue * Multiplier + Adder + BaseValue
Calibration constants are typically updated each year or more frequently when stations and sensors are maintained.
The NovaStar database stores the raw value to be a floating point number. However, the value may be an integer, measured using sensor units.
If data are ingested as scaled value, the corresponding raw value is back-calculated.
If no conversion is required, the adder is set to 1 and the multiplier and base to 0.
Calibration Data Best Practices
- Calibrations are specified for a point and are not shared. However, default configuration information can be specified in point type data and can be used for simple cases.
- Make sure that one calibration is defined with an effective time before the first data report. A calibration will be used for all data including and after the effective time, until the next calibration and effective time are encountered.
- If data are edited in the database, rerun calibrations to ensure that derived data are recalculated:
- raw data values will be calculated if scaled values were edited
- scaled values will be calculated if raw values were edited
- If appropriate, add notes to the point description to describe major changes in configuration.