Roles and Applications for Hydro Ontologies

Joshua Lieberman Harvard Center for Geographic Analysis

jlieberman@fas.harvard.edu

What is an ontology? What is linked data? Information graph perspectives Ontology roles Hydrography / hydrology ontologies Putting hydro ontologies to work

What is an ontology?

Specification of a conceptualization* Theory within universe of discourse Entities and their relationships Semantics? Predicate logic Tbox & Abox Axioms

*Tom Gruber

!Linked (Open) Data

Web of data" RDF Triples URL URI's SPARQL Distribution and

dereferencing

Ontology

Information Graph Perspectives

Logical relationships Spatial relationships Physical networks Distribution networks

Roles for Ontologies

Classification Discovery Data integration Data provenance / orchestration Knowledge entailment Application logic Model configuration

Hydro Ontology Uses & Cases Classification of monitoring data according to hydro

conventions Physical / virtual contracted nodes points of reporting

(Aus) Linking and 2N mediation between hydro models Discovery and extraction of hydro features and data

through links Map data as a feature framework Feature framework for place names Flow & containment framework for water-related

modeling Flow framework for flood prediction (e.g. NFIE and 2.7M

virtual gauges)

Links and linked data Ontologies identify entities with globally

unique URIs Linked data dereferences URIs to URLs with

useful results. A proper hash of it

http://usgs.gov/nhd/flowline.349530# Fair representation (.xml, .json) Give it a REST API Final resolution (SPARQL and otherwise)\ Identity questions (Reachcode,

Permanent_ID, COM_ID, etc.)

Work in Hydro Ontologies

CUAHSI Ontology HY_Features Ontology CEGIS NHD Ontology OGC Testbed 11 Ontology Hydro Ontology Wet-dry Ontology

Semantic Sensor Networks PROV-O

Upper ontologies vs Patterns

used

endedAtTime

wasAssociatedWithactedOnBehalfOf

wasGeneratedBy

wasAttributedTo

wasDerivedFrom

wasInformedBy

Activity

Entity

Agent

xsd:dateTime

startedAtTime

xsd:dateTime

Semantic Sensor Networks PROV-O O&M

Upper ontologies vs Patterns

CUAHSI HIS Ontology

NHD Ontology

HY Features Ontology

HydroOntology

60

Figure 4.8 Class Hierarchy of HydroOntology

N. Vijayasankaran MS Thesis

HydroOntology

N. Vijayasankaran MS Thesis

61

A river or stream has a source or headwaters, a mouth which is a discharge point, and

zero to many tributaries. These relations express that a river and other hydrologic

features are geographically connected and contribute inflow or outflow of water to each

other. Sources of a river can be other hydrological features such as a Spring, River or

Lake. Similarly the terminal feature of a river can be another river, a lake, a Bay or an

Ocean. By specifying object properties between feature classes, such as River hasSource

Spring , River hasMouth Lake , River hasTributary River, it becomes possible to query

and make inferences from the hydro-gazetteer on hydrological relationship between

proper named features.

Figure 4.9 Hydrographic Relationships of River to Other Feature Types

Sample SPARQL Query

SPARQL query below retrieves all the streams that are part of the main stem along with dam locations.

SELECT ?name ?damname WHERE {?s gaz:gnisName 'Androscoggin River'. ?s ho:hasMainStem ?feature. ?feature gaz:gnisName ?name. ?feature ho:hasDam ?damfeature. ?damfeature gaz:gnisName ?damname}

Surface Water Features Fig. 2. An illustration of how surface water features can be described as instances of classes

defined in the Dry (left) and Wet (right) modules of the Surface Water pattern.

Surface Water Patterns Fig. 1. Surface Water patterns Dry and Wet

module classes (brown/blue) and properties (grey).

Activities

Ontology alignment metrics Aligned flow network ontology pattern Adaptation of hydraulic analysis techniques for channel

delineation Contouring approach to surface network resolution Social nexus approach to placename relationships Model for generation and management of

dereferenceable feature URIs Extraction of entities from NHD and NHDPlus to

populate and exercise knowledgebase Ontology-driven applications Further TNM Linked Data work in OGC Testbed 12

REST: Follow the Links Get a bunch of features

http://www.pvretano.com/cubewerx/cubeserv/default/wfs/2.5.0/ows11/wwAccess Get a specific feature

http://www.pvretano.com/cubewerx/cubeserv/default/wfs/2.5.0/ows11/wwAccess?featureId=CWFID.WWACCESS.0.33678.BA8CA5E7ADF849B81F20020000

Get a feature in GML http://www.pvretano.com/cubewerx/cubeserv/default/wfs/2.5.0/ows11/wwAccess?featureId=CWFID.WWACCESS.0.33678.BA8CA5E7ADF849B81F20020000&f=application%2Fgml%2Bxml%3B+version=3.1

Follow the link with rel="alternate" and type="application/vnd.geo+json" to get JSON http://www.pvretano.com/cubewerx/cubeserv/default/wfs/2.5.0/ows11/wwAccess/CWFID.WWACCESS.0.33678.BA8CA5E7ADF849B81F20020000?outputFormat=application%2Fvnd.geo%2Bjson

Query Bbox http://www.pvretano.com/cubewerx/cubeserv/default/wfs/2.5.0/ows11/wwAccess?bbox=-43.4380,172.6489,-43.3947,172.7108&css=http://www.pvretano.com/cubewerx/css/ows11_table.css

Follow the link with rel="collection" to get the featuretype, default 10 features. http://www.pvretano.com/cubewerx/cubeserv/default/wfs/2.5.0/ows11/wwAccess?css=http://www.pvretano.com/cubewerx/css/ows11_table.css

Follow the link with rel="service" to get the service document of the service. http://www.pvretano.com/cubewerx/cubeserv/default/wfs/2.5.0/ows11

Copyright 2015 Open GeospaBal ConsorBum