NovaStar Data Model / Rating
NovaStar data records contain raw, scale, and optionally rated values (e.g., discharge computed from water level). A rating indicates how to convert a scaled value to a rated value.
Overview
Each point's data can have up to 5 rated values computed from the scaled value. Rating computations include:
- Conversion using a rating equation or table (e.g., water level to discharge)
- Built-in calculations (e.g., precipitation calculations):
- Precipitation increment calculated from accumulated value.
- Precipitation storm total (from increments).
- Precipitation seasonal total (from increments).
Many points will only use raw and scaled values without ratings.
Rating Tables
A rating equation or table is typically used for:
- Water level (scaled value) to discharge (rated value).
- Reservoir level/area/volume/seep estimates.
Rating equations and tables have effective time to allow ratings to be updated based on field measurements such as stream surveys. For example, a rating table may be updated when the channel configuration changes.
Ratings that involve elevation must be configured to use an elevation datum such as mean sea level (MSL) or other standard datum, or a surveyed local datum such as a survey marker.
It is theoretically possible to share rating tables between points. However, practically, rating tables are often unique to a point. The NovaStar design uses "rating assignment" (rating assign) data to relate a point to a list of rating equation or table, with the effective time used to sort the ratings in chronological order for processing.
The rating assign "computation type" indicates the type of calculation to perform to convert the scaled value into rated value. The SHEF parameter code indicates the data type for rated values.
NovaStar provides a number of options to convert input data (point scaled value) to output.
Point Configuration for Rated Values
The data records for points can contain up to 5 rated values, which are identified using a "rating number" (1 - 5). Many point data types do not use the ratings because only the original value is of significance for the report time.
Interval data (e.g., 1-hour precipitation total or 15-minute mean streamflow) can be computed by web services and NovaStar command line tools and don't typically need to be stored in the database unless the original data from a data logger, import, or equation are interval data.
It is desirable that the rated values follow the same order and conventions, regardless of how data are inserted in the database:
- for precipitation, rated values should be the incremental precipitation, season (or annual) total, and storm total, regardless of whether accumulated precipitation or precipitation increments were loaded
- for streamflow, the scaled value should be water level and first rated value should be the discharge computed from the water level
This allows third-party applications to use the rated value numbers consistently
without having to know how the data were inserted.
For more failsafe data use without relying on the rating number,
use data web services, which use the rated value data type string (e.g., DischargeRiver
) rather than the numerical position.
The following table summarizes the recommended point configuration for rated values.
Point Configuration for Rated Values
Point Data Type | Rated Value Configuration |
---|---|
Accumulated precipitation (e.g., ALERT/ALERT2)
|
|
Incremental precipitation (e.g., individual bucket tips or import of small-interval increments such as NWS MADIS ASOS data
|
|
Water level |
|
Rating Table Data Best Practices
- If a rating table is used, make sure that one rating is defined with a rating assign effective time before the first data report. The rating assign (and associated rating table, etc.) will be used until a later effective time is encountered.
- If rating data are edited in the database, rerun ratings to ensure that derived data are recalculated.
- Add notes to the rating table description to describe the source of the rating table data.