Enhancing Data Quality Assessment Capabilities by Providing Unique, Authoritative, Discoverable, Referenceable Sensor Model Descriptions

With observational data becoming widely available, researchers struggle to ﬁnd information enabling assessment for its reliable use. A small ﬁrst-step toward enabling data quality assessment of observational data is to associate the data with the sensor used to make the observations and to have the sensor description machine-harvestable. In the latest additions to the X-DOMES (Cross-Domain Observational Metadata for Enviromental Sensing) toolset, we have created targeted editors for creating SensorML documents to describe sensor models. The team has adjusted its delivery to enable integration of the X-DOMES content with the GEOCODES (JSON-LD/schema.org) EarthCube project. At our poster-session, we will highlight the new changes and capabilities and demonstrate the use of new X-DOMES tools.


PROJECT GOALS
We seek to enable and encourage the creation of process descriptions that are needed to assess data quality for archival and reuse.Sensor metadata are to be made available using community-adopted standards (OGC/W3C) to assure FAIR data practices in our earth observations and its associated metadata.
The team has created tools for the non-expert to create and manage SensorML documents.The tools enable the creation of machine-actionable content.
X-DOMES implementation will:

•
Help large observational data producers automate and manage sensor and operational provenance • Encourage small federally funded data providers to describe sensor data in ways that meet agency requirements for data management

•
Facilitate creation of and access to common content using standards-based production of interoperable sensor documentation by environmental sensor manufacturers

•
Enable data aggregation centers to build relationships across domains for integration of sensor-based observations

•
Provide the ability to assess data quality and automate quality control, based upon manufacturers' descriptions of sensor provenance

•
Generate registries for sensor and deployment metadata that can be utilized by building blocks of a layered architectural cyberinfrastructure [1] OGC: SENSORML SensorML can describe how an observation came to be, enabling a better understanding of the data.It is about "systems" that can be applied across-domains.
A system description is composed of input (observable property) and output (observation) and the process by which the output is created.It describes the sensor, the parameters, and the processes used in creating observations.Each sensor has characteristics and capabilities that can affect the interpretation and the quality of the observation.Each action that affects the data and data quality should be described by the agent having the knowledge of the action.
The first step in describing how an observation was made is to define the sensor used.A sensor manufacturer can describe a sensor model (Original equipment Manufacturer, OEM) and then referenced the OEM in a description of the as-built, asconfigured Instance sensor.The image below indicates how the documents are inter-related and how they can be utilized.

Our team has created two versions of an online SensorML editor/viewer.
The BASIC editor is easier to navigate but limited in its ability to incorporate full descriptions and links to terms.
The full-featured editor requires some functional knowledge of how to describe your terms.There is a video available on the ESIP Community Page (https://www.esipfed.org/earthcube-xdomes).
. Once the content is created.It can be registered and maintained by the XDOMES SensorML Registry and Repository (SRR).This act provides unique identifiers that can be used in associating the sensor to data and publications.
The tools are ready for stakeholders to begin creating and registering the descriptions of sensor models.And to describe a sensor Instance (as built sensor) that provides a unique description of a particular sensor.These documents can be registered -thereby creating a unique ID that can be referenced in data systems and publications.
TO GET STARTED: Check out and save a file from the SRR, then load it into the Basic Editor.Then start describing your sensor!OGC+W3C: SENSOR WEB Above shows the tasks, tools and interrelationship of the resources.
To facilitate the creation of content and to register the content, the XDOMES team has created: Ontology Registry and Repository (ORR) where a user can link to existing terms in community adopted ontologies or create their own set of terms to describe their content.(Instructions are on the ESIP Community Page (https://www.esipfed.org/earthcubexdomes).)

PROJECT TEAM AND ACKNOWLEDGEMENTS
Janet Fredericks, as the principal investigator (PI) on the Q2O project (http://q2o.whoi.edu),led the development of the quality assessment model [4] that has been used by X-DOMES.She also brings with her many years of research experience, both as an operational oceanographer and a systems programmer.She managed the cabled observatory called the Martha's Vineyard Coastal Observatory, which hosts several research projects each year and has been serving a suite of real-time sensor data since 2001.She served as a liaison to the Inter-Agency Ocean Observation Committee DMAC-ST and also the U.S. IOOS Quality in Real-Time Oceanographic Data Board of Advisors and a participant of the EU-Australia-US Ocean Data Interoperability Platform project.She was involved in EarthCube in its initial phase as a member of the Brokering, Data Access and Semantics & Ontologies community groups and a funded participant in the Layered Architecture Concept Award and participated in the Technical Architecture Committee and the Gap Analysis Working Group of the EarthCube test governance.She has also served two years on the EarthCube Leadership Council.
Mr. Felimon Gayanilo is the Systems Architect for two major scientific data portals in the Gulf of Mexico.The first is the Gulf of Mexico Coastal and Ocean Observing System (GCOOS) that is nested in a National Backbone of coastal observations to aggregate and disseminate the region's near real-time oceanographic data in OGC Sensor Observation Service (SOS) to facilitate interoperability among coastal and ocean observing sensors.The second is the Gulf of Mexico Research Initiative Data and Information Cooperative (GRIIDC), a 10-year multi-disciplinary and multi-institutional research effort in response to the 2010 Deep Horizon oil spill disaster.
Dr. Mike Botts is the author of SensorML and has served as the chair of the OGC® SWE Domain Working Group since its conception.He received the 2008 Gardels Medal for his role in leading the SWE standards activities in OGC®.He is also the lead for development of the current Advanced SensorML Editor and PrettyView, the Space-Time Toolkit visualization package and a variety of open-source libraries in support of SWE.He is currently managing a project to develop an open-source SensorHub (http://opensensorhub.org) to support easy deployment of sensors with immediate access and tasking through SWE 2.0 standards.He also was a Co-PI in the Q2O project.Dr. Botts was an elected member of the OGC Architecture Board (2008)(2009)(2010)(2011)(2012)(2013)(2014).
John Graybeal co-founded the Marine Metadata Interoperability Project in 2004, and continues to serve as the Project Lead.As part of his role he co-hosted several workshops, including the 2006 Sensor Metadata Interoperability workshop.He led MBARI's development of the Shore Side Data System, and guided Data Browser features for Marinexplore/PlanetOS.He served on the W3C's Semantic Sensor Network Working group, served as the co-chair of the ESIP Federation's Attribute Conventions for Data Discovery, and wrote the NetCDF Climate and Forecast Conventions Frequently Asked Questions.He continues to develop and refine vocabularies served by MMI's Ontology Registry and Repository and has sheparded its adoption as a community resource through ESIP (COR) (http://cor.esipfed.org/).Dr. Carlos Rueda (MBARI) has been the technical lead for the MMI ORR software used for XDOMES, and also used for the MMI Ontology Registry and Repository and ESIP Community Ontology Repository.He coordinated an international group 6/23/20, 11:00 AM toward the development of a marine device ontology (elements of which were adopted by the W3C Semantic Sensor Network ontology effort) and has been the main developer of the MMI ORR system.He assists IOOS, OOI, ICAN, and other communities in the development of controlled vocabularies and has co-hosted international technical meetings with ontology repository developers, promoting the need to define inter-repository standard interfaces.He has participated in the DataONE Semantics and Integration Working group and the Open Ontology Repository initiative.
Dr. Krzysztof Janowicz is an professor for Geographic Information Science at the Geography Department of the University of California, Santa Barbara, USA as well as one of the two Editors-in-Chief of the Semantic Web journal.He is also the community leader of the 52°North Initiative for Geospatial Open Source Software GmbH semantics community that develops open source solutions that bridge OGC's Geo-Web and the Semantic Web.Janowicz is a member of the W3C SSN-XG that developed the Semantic Sensor Network ontology (SSN), and was responsible for the development of the Stimulus-Sensor-Observation ontology design pattern (SSO) that forms the core of the SSN ontology.Besides ontologies, Janowicz has developed software and specifications for sensor mediation, Semantic Enablement of Spatial Data Infrastructures, and Restful Linked Data proxies for the OGC Sensor Observation Service.He published large Linked Data sets such as the ADL gazetteer and tools for their exploration.He was responsible for the alignment of the X-DOMES SensorML work with the W3C/SOSA efforts.