1 CEOS WGISS project: GEOSS Reference Model for the Use of Satellite Data in Disaster Management and Risk Assessment John Evans - GST, Inc. Karen Moe - NASA Earth Science Technology Office GEOSS Future Products Workshop March 2013

2 Context

3 Problem Statement International disaster management involves: Many activities by many players Many ad hoc arrangements => Limited effectiveness, efficiency New suppliers: unclear how to contribute data & services New users: unclear how to tap into these data & services Unclear what resources are Shared Missing Interdependent Isolated Need to establish partnerships, standards, shared vocabulary, etc., in advance of disaster events Need a precise, common understanding of processes, information & computing resources, and user needs

4 Desired outcomes Clear, shared understanding Information systems & services: components, roles, relationships Effective, efficient, collaborative processes & systems Streamlined, automated access to data, services Scope: disaster mitigation, warning, response, & recovery Ongoing activities linked to overall enterprise Proof-of-concept prototypes, GEOSS AIP, CEOS DRM Shortfalls, gaps, redundancies identified Insights from (and relevant to) practitioner experience

5 Approach Describe satellite support to disaster management using a formal architecture framework Classes, relationships, interdependencies 3 Viewpoints: Enterprise, Information, Computation Analyze experiences with satellite support to disaster management Int’l Charter; Disaster Management Constellation; others East Japan earthquake (2011); NASA / Namibia Sensor Web Flood Pilot; Sichuan (China) earthquake (2008) Capture lessons learned; Recommend standards; Identify building blocks for sustainable capability Guidance for GEOSS, CEOS activities Disaster portals ( e.g., CEOS, UN-SPIDER, others) CEOS Disaster Risk Management (DRM) initiative

6 1 Based on CEOS / GEO DI report, “Use of Satellites for Risk Management” (2008) Scope, purpose, structure Scope & purpose based on GEO Task DI-01 GEOSS Strategic Targets CEOS WGISS charter Consistent with GEOSS principles System of Systems Data Sharing Principles Interoperability Arrangements Disaster types 1 Flooding Earthquakes Volcanoes Drought Windstorms Landslides Wildfires Tsunamis Lifecycle phases 1 Mitigation Warning Response Recovery Enterprise Viewpoint

7 Information content & semantics GEOSS AIP Architecture concepts Spatial referencing Feature Model Data Quality / Provenance Data Policies / Licensing etc. Observations by disaster type & phase – based on Work Plan (GEO, 2012) Work Plan Critical Earth Observations Priorities (GEO report, 2010) Critical Earth Observations Priorities GEOSS 10-Year Implementation Plan Reference Document (GEO, 2005) GEOSS 10-Year Implementation Plan Reference Document Use of Satellite Data for Risk Management (CEOS/GEO report, 2008) Use of Satellite Data for Risk Management Metadata Finding relevant data Assessing fitness for use Georeferencing Semantics Information Viewpoint

8 Information content & semantics Example: Information needs for Wildfires Information Viewpoint

9 Information content & semantics Practitioner insights Cross-cutting needs: Frequent, high-resolution observations Esp. for earthquake, flood response & recovery Basemaps – e.g., Digital terrain models – Water boundaries – Ground control points Common data operations Preprocessing (e.g., decoding, georeferencing, atmospheric correction – “Level 1”) Analysis & Interpretation (incl. feature extraction) Product creation (incl. “image pyramids” for visualization) Information Viewpoint

10 Computation and Services Generic service types (from AIP-5 architecture) : Catalog Registration & Search Portrayal / Display / Styling Data Access & Ordering Processing algorithms Sensor access & control User management Disaster-specific service types: Event detection Sensor tasking Data Analysis / Interpretation Modeling / Prediction Namibia Flood Pilot Sensor Web Concept Computation Viewpoint

11 Computation and Services Practitioner insights Cross-cutting needs: Near-Real-Time data access / delivery Cross-community interoperability Ease of use; ease of operation / maintenance “Last mile” to end-users (incl. telecomm. infrastructure) Web Services & other services Broadcast / push (LDM, GeoNetCast) Physical media delivery Computation Viewpoint

12 Functions involved in Satellite Data Support to Disaster Management User-generated postings Routine / Global Monitoring 1. Event detection Routine / Global model outputs 4. Modeling Task sensors In situ Remote Acquire data 3. Data acquisition Archive Visualize Ongoing updates Alerts / Notices Publish product 6. Dissemination Preprocess Analyze 5. Analysis Interpret Decision Analysis: Act / Plan 2. Situational Awareness Gather / Assimilate Information Local Forecast Hindcast Nowcast

13 Phases of Disaster Management Capability System initiation vs. steady-state operations Initiation: Identify inputs for event detection; event triggers Choose indicators for situational awareness ( e.g., flood extent) Define modeling elements ( e.g., regional flood model) Develop workflows and data flows (for processing and delivery) Define automation goals ( e.g., subscriptions, custom products) Steady state: Monitor data streams, detect events & trigger workflows Track key indicators Task sensors; Acquire data Run models (hindcast, nowcast, forecast) Analyze and disseminate products

14 Preliminary recommendations Need coordination mechanisms for all phases of the disaster lifecycle Mitigation - Warning - Response - Recovery Need a services infrastructure to streamline access Near-real-time services On-demand, user-customizable products Specialized end-user tools (apply generic data & services) Meeting diverse user needs (JPEG images vs. quantitative data grids) Need open, well-defined interfaces Data access Data processing & interpretation Modeling Sensor tasking 1/14/

15 Preliminary recommendations Metadata describing fitness for use is crucial Operational decisions require knowing data quality Can’t just filter out all imperfect data … As are tools for finding / ranking / filtering data from multiple sources Based on spatial & temporal coverage; resolution; cloud cover; Based on expected use ( e.g., wildfire detection) 2-way collaboration is increasingly important Providers may also be co-analysts, who may also be end users Need broader, more accessible data access & data sharing Ease restrictions on redistribution Crucial to providing frequent, high-resolution observations

16 State of the art Service-oriented processes, incl. a few kinds of Service interoperability Sensor tasking & services Customized, on-demand products User-generated / crowd-sourced data Metadata that begin to support Catalog search Assessing fitness for use Workflow automation Clear semantics

17 Near-future directions Robust services infrastructure Simple, predictable service interoperability Widespread, standards-based sensor access & tasking Simple access to customized, on-demand products Standards / Infrastructures for user-generated data Metadata that can fully support Reliable data discovery and filtering Workflow automation => high-level user interfaces Semantic linking across providers and communities

18

19 Framework: ISO/IEC Reference Model of Open Distributed Processing (RM-ODP) Enterprise viewpoint: the purpose, scope, and policies for the system. Often articulated by means of use cases. Information viewpoint: the semantics of the information and the information processing performed. Computation viewpoint: the functional decomposition of the system into objects interacting at interfaces. RM-ODP is the basis for GEOSS Architecture Implementation Pilot (AIP), E.U. ORCHESTRA, OGC Reference Model, and others

20 Framework: ISO/IEC Reference Model of Open Distributed Processing (RM-ODP)

21 Enterprise view: purpose / scope GEO Task DI-01, “Informing Risk Management and Disaster Reduction” seeks to achieve the following: More timely dissemination of information from globally- coordinated systems for hazard monitoring, prediction, risk assessment, early warning, mitigation, and response. Multi-hazard and/or end-to-end approaches to disaster risk reduction, preparedness, and response. Support for the Hyogo Framework for Action Improved use of observations in policies, decisions and actions associated with disaster preparedness and mitigation. More effective access to observations to facilitate disaster warning, response and recovery. Increased communication and coordination between national, regional and global communities. Improved disaster response through delivery of space-based data, via the International Charter on Space and Major Disasters.

22 Enterprise view: purpose / scope GEO DI-01 focus areas: Provide support to operational systems Enable and inform risk and vulnerability analyses Conduct regional end-to-end pilots with a focus on building institutional relationships Conduct gap analyses in order to identify missing data, system gaps, and capacity gaps GEO DI-01 components: Disaster Management Systems Geohazards Monitoring, Alert, and Risk Assessment Tsunami Early Warning and Hazard Assessment Global Wildland Fire Information System Regional End-to-End Pilots GEO DI-01 implementation Resources

23 Enterprise view: Stakeholders Often mentioned; seldom characterized or enumerated Case studies will shed light on this from practitioner perspectives GEOSS AIP-3 (01/2010): “targeted or supported” communities National agencies concerned with disaster management, meteorology, hydrology, and emergency response, and their supporting providers of data, services, research, and analysis CEOS Strategic Implementation Team (SIT) and WGISS GEOSS' DI (=> DI-01) Task UN-SPIDER GEOSS AIP-3 Disaster Management reference scenario: Initiators (trigger and coordinate the disaster response) Actuators (respond to disaster – e.g., regional civil protection, insurance companies, NGOs) Processors (provide raw data or derived information) Coordinators (facilitate interactions among the other actors)

24 Enterprise view: Processes Information support activities (from GEOSS AIP-5 architecture)

25 GEOSS Data Sharing Principles System of Systems Independently operated systems contributed to (also) serve shared purposes Data Sharing Principles Full and open exchange of data Minimum delay and cost Support to research or education at zero or marginal cost Interoperability Arrangements Industry & international interface standards Adopted by the GEO Standards and Interoperability Forum (SIF) Maintained in the GEO Standards Registry

26 Enterprise view: points of comparison Example: International Charter Supply space-based data to relief efforts in the aftermath of major disasters Differences in scope w/ GA.4.D enterprise: Support disaster relief – not research, prevention, etc. Supply data products – not original data or end-user services

27 Enterprise view: points of comparison Example: GeoHazard Supersites Open access to data for 16 seismically active sites around the world Spaceborne SAR; GPS deformation measures; earthquake observations Differences in scope with GA.4.D enterprise: Seismic risks only – not floods, storms, etc. Emphasis is on research – not operations (so far)