There is a large diversity in the amount and type of water data collected across the state and needed to inform water-related decisions. Data could be about
surface water conditions (e.g., evaporation, soil moisture, precipitation, ecosystem health, and location of reservoirs/rivers/conveyances);
ground water conditions (e.g., groundwater levels, locations and depths of wells, aquifer parameters, and geological formations); and
related management processes (e.g., permits for use (type and quantity), measured diversions/extractions, amount in storage, water returns/injections, and water quality).
For all these datasets, we need to capture relevant information (i.e., specific measurements) as well as associated metadata (i.e., describes and gives information about the specific measurements).
Generally, every agency has their own internal practices and standards but we require a common understanding to ensure data is being accurately represented within the central platform.
We adopted a SensorThings (ST) data model to develop a common “language” around the different types of water data. ST is a service focused on data streaming that supports integration of metadata within the data structure.
The basic elements of SensorThings are:
Location - physical location of the observations. Can be a point or polygon. Captured in a geoJSON format.
Thing - A collection of one or more sensors/an abstract description of what is being measured e.g Water Levels
Sensor - The sensor doing the measurement
Datastream - A group of observations of an observed property by a sensor for a thing at a given location
ObservedProperty - Description of what is being observed
Observations - The actual time vs intensity data
Kyle Onda (Unlicensed) from the Internet of Water provided us an overview of the ST framework on . Check out the recording from his talk here. A copy of his slides are available
Slides:
0 Comments