Use of WPS for Earth Observation 4 th

Use of WPS for Earth Observation 4 th GEOSS Science and Technology Stakeholder Workshop 25 March 2015 Sheraton Norfolk Waterside Hotel

Agenda • Intro WPS – Basic usage – Chaining • Use of WPS in the Architecture Implementation Pilot Phase 8 – Use of chained WPS for Data Quality Assurance • (Mike Jackson, Didier G. Leibovici, Sam Meek - COBWEB) • Big Data Analysis – WPS and HADOOP

Intro WPS • Principle – Similar to Unix filter • stdin > processing > stdout • f(x) = y • WPS is like a “Unix filter” on the Web

Intro WPS • Behaves like the other OGC Web Services (WMS, WFS, WCS, CSW, …) Diagram by 52 North

Get. Capabilities Diagram by Geo-Solutions. it

Describe. Process Diagram by Geo-Solutions. it

Execute Diagram by Geo-Solutions. it

WPS Chaining • Remember the Unix filter from slide 3? – Connect stdout to stdin using | • ls -l | grep ”Mar” • ls -l | grep ”Mar" | sort +4 n | more • WPS’s can also be chained! – So you can run more complex analysis (model)

Using WPS • Use of WPS in the Architecture Implementation Pilot Phase 8 (AIP-8) – Use of chained WPS for Data Quality Assurance • (Mike Jackson, Didier G. Leibovici, Sam Meek – COBWEB)

AIP Participants using WPS SBA Organisation JAXA Connecting. GEO Cloud. Sigma SDI 4 Apps EU BON FOODIE Smart. Open. Data GEOSS Water Services Chile FSU-Jena NATO STO-CMRE CIESIN COBWEB CNR Storm. Center NASA Ben Burford GMU INCOSE CITI-SENSE He Cl We Dm Ag Technical Wa X Bd En Other Auth Proces sing LD Cloud X ? X X WS in GCI X X X Ocean X X X X X X X X Air. Q X X X X X X X X Portal App X X Sen sor X X ? X X X X X X X CB & BP X X X X X

Example • Use of chained WPS for Data Quality Assurance – (Mike Jackson, Didier G. Leibovici, Sam Meek – COBWEB) • Context: – VGI – Citizen Science – Citizens Observing in Biospheres

Quality assuring data through atomic tests – 7 Pillars

System Architecture

User interface

Example workflow: Ash Dieback risk in the Dyfi Biosphere

Reversal • f(x) = y • We download the Earth (x) to many times a day – Data lives happily in the cloud – Bandwidth and Processing power is infinite in the cloud(s) (almost, sort of) • Upload f() to x, and download y! – F() is relatively small and y is expected to be small • WPS-Transactional! – Upload the f() to the Web Service

Big Data Analysis • WPS with HADOOP, SPARK, … • http: //www. opengeospatial. org/blog/1866 • Further investigated in OGC Test. Beds and AIP-8

Thank you! • Questions? • bdelathouwer@opengeospatial. org