Development of an Information Exchange format for the Observations Data Model version 2 using OGC Observations and Measures

Abstract:

The Observations Data Model v1 (ODMv1) schema has been utilized of the basis hydrologic cyberinfrastructures include the CUAHSI HIS. The first version of ODM focused on timeseries, and ultimately led the development of OGC “WaterML2 Part 1: Timeseries”, which is being proposed to be developed into OGC TimeseriesML.Our team has developed an ODMv2 model to address ODMv1 shortcomings, and to encompass a wider community of spatially discrete, feature-based earth observations. The development process included collecting requirements from several existing Earth Observations data systems: HIS,

CZOData, IEDA and EarthChem system, and IOOS. We developed ODM2 as a set of core entities with additional extensioncomponents that can be utilized. These extensions include for shared functionality (e.g. data quality, provenance), as well as specific use cases (e.g. laboratory analysis, equipment).

Initially, we closely followed the Observations and Measures (ISO19156) concept model. After prototyping and reviewing the requirements, we extended the ODMv2 conceptual model to include entities to document ancillary acts that do not always produce a result. Differing from O&M where acts are expected to produce a result. ODMv2 includes the core concept of an “Action” which encapsulates activities or actions associated that are performed in the process of making an observation, but may not produce a result. Actions, such as a sample analysis, that observe a property and produce a result are equivalent to O&M observation. But in many use cases, many actions have no resulting observation. Examples of such actions are a site visit or sample preparation (splitting of a sample). These actions are part of a chain of actions, iwhich produce the final observation.

Overall the ODMv2 generally follows the O&M conceptual model. The nearly final ODMv2 includes a core and extensions. The core entities include actions, feature actions (observations), datasets (groupings), methods (procedures), sampling features and variables (observed properties). Results entities are separated to handle the diversity of possibilities: time series, categorical, and 6 others. Extensions handle requirements for equipment, laboratory analyses, provenance, data quality, and controlled vocabularies.