Tsunami Warning Center Operations System TOPS Critical Design

Tsunami Warning Center Operations System (TOPS) Critical Design Review (CDR) February 20, 2014 1

Agenda • • • Introduction Overview Preliminary Design Review (PDR) findings Critical Design Review (CDR) topics Wrap Up 2

Introduction • CDR purpose: – Provides detailed system architecture for hardware, software, and data. – Allows the Government to achieve confidence that the design of the tsunami warning system will meet the objectives, functional requirements, and technical requirements of the project and is in conformance with the National Weather Service (NWS) enterprise solution (AWIPS II). • Presenters: – Brian Donnelly, Project Manager – Frank Griffith, Lead Architect – Sid Hellman, Lead Scientist 3

Introduction • Primary ERT team members: – Raytheon, Intelligence and Information Systems (IIS) – Instrumental Software Technologies, Inc. (ISTI) • Extended team members: – gempa Gmb. H – Anthony Lomax, ALomax Scientific 4

Overview Agile development sprints management TOPS is built in 13 one month sprints with demonstrations every three months Agile sprints management in the Redmine web portal enables tracking of all project activity Redmine development tasks Gantt chart 5

Overview Collaboration through Redmine • Increase the efficiency of collaboration through web-based sharing using the project portal • NWS can appoint someone (or a set of individuals) to have web portal Redmine access to facilitate sharing • Example: A developer needs clarification or test data for a requirement: – NWS can add a comment to the referenced requirement or attach data files – Redmine will auto send an e-mail to the developer when the update occurs 6

PDR FINDINGS 7

6. 1. 1 PDR Findings Building off original PDR design concept TOPS builds on the functional SOA concepts described in the PDR Common Plug -in Based Ingest Pattern Standard Agile High Performance Framework Ingest Processing Seismic Ingest Sea Level Ingest Consolidates using plug-in Extendable CAVE Leverages New NWS Hazard Services Tsunami Forecast GIS Display Product Generation Seismic Processing Precomputed GIS Overlays Text Based Sea Level Processing Dynamic Views Internet Based Flexible Adapter Design BUS Event Driven Interfaces DBMS There were no findings from the PDR Central Historic & Real-Time Storage Data available to all components 8

6. 1. 1 PDR Findings Building off original PDR design concept Design refinements since the PDR Refinements Description CMT Integration Move the NEIC CMT into Seis. Com. P 3 and have a Quake. ML message generated with the CMT results to communicate with EDEX Seismic Alarming Add seismic alarming generation to Seis. Com. P 3 and add a new message queue to communicate alarms to EDEX/CAVE using the CAP protocol Sea Level Processing Move processing into CAVE (from the PDR concept of having it in EDEX) to facilitate the dynamic interactions that are available today in “Tide. View” Tsunami Forecasts Models Minimize the changes to the existing forecast models so that they can run in their existing environments. Build the adapters so that the data comes from the DBMS but gets put into existing file interfaces for running. Product Generation Reuse only the AWIPS Hazard Services Product Generation plug-ins and build new dialogs for the control. There were no findings from the PDR 9

CDR TOPICS 10

6. 1. 2 CDR Logical Data Model Database tables implemented in Postgre. SQL Prototype Implemented Phase 2 - To Be Implemented SNMP Tool Historical Tables Status entries New… Seismic waveform data Historical earthquake parameter info Monitoring and system status messages Sea level data Historical tsunami parameter info Error messages Seismic station metadata Historical tsunami marigram data Sea level station metadata GIS overlay info Event parameter data Tsunami travel-time info Tsunami Threat Data Wave-height forecasting info EDEX Log 4 j log files Log entry 11

6. 1. 3 CDR Physical Data Model Seismic and station data Postgre. SQL Metadata Tables Seismic. Record {reftime, rangeend, station_id, samplerate} Database tables auto-generated by plugin Metadata Tables and Storage Defined by Annotations Example: Seismic. Record. java Column Name Type Station Seismic. Station start. Sequence Int end. Sequence Int sample. Rate double Comment Data. URI (1) Seismic waveform data seismic_station {networkcode, stationid, locationid, channelid} Column Name HDF 5 Data Store Data. URI Type Comment Id Int Auto station. Id String Data. URI (0) network. Code String Data. URI (1) location. Id String Data. URI (2) channel. Id String Data. URI (3) Latitude Double Longnitude Double Elevation Double Description String Seismic station metadata 12

6. 1. 3 CDR Physical Data Model Seismic event data Event parameter data Postgre. SQL Metadata Tables Seismic. Event {event. Identifier, version} Column Name Type Comment Agency String (Unknown) Data. URI (1) Author String (Unknown) Data. URI (2) event. Type Event. Type (Automatic, Manual) Data. URI (3) event. Identifier String Data. URI (4) event. Magnitude Version Int event. Location Picks Set<Pick> Data. URI (5) Extend to CMT data Location{} Column Name Tsunami travel-time info Type Comment Magnitude{reftime, agency, author, magnitude, type, latitude, longitude} Column Name Type Comment Latitude Double Data. URI (4) Longitude Double Data. URI (5) Magnitude Double Data. URI (4) depth Double Data. URI (6) Type String Data. URI (5) 13

6. 1. 3 CDR Physical Data Model Sea level data Postgre. SQL Metadata Tables Sea. Level. Data {} Column Name Type site. ID int sea. Level float datum. Offset Float time. Stamp String ep. Time long Comment Data. URI HDF 5 Data Store 14

6. 1. 3 CDR Physical Data Model Sea level station data Postgre. SQL Metadata Tables Sea level station metadata Column names from Tide. View Sea. Level. Stations{? } Column Name Type Comment Column Name Type Site. Name String Recording. Agency String Site. Abbrev String Sensor. Type String WMOCode String Ocean String GOESID String Comments String Site. Operator String Data. Refresh Int Site 0 Abbrev String Data. Refresh. Units String Site. Lat Double Site. Lon Double Water. Level. Units String Transmission. Type String Sample. Rate. Units String Reference. Datum String Comment 15

6. 1. 3 CDR Physical Data Model Tsunami threat data Column names from Tide. View. h Tsunami Threat Data Postgre. SQL Metadata Tables Tusnami. Event{? } Column Name Type Status Int Site String ETATime Long Fcst. Amp Double Fcst. Draw. Dn Double Fcst. Pk. Cur Double Obs. Amp Double Period Double Obs. Time Long Eobs. Time Long Arr. Time Long Save. File Boolean Has. Arrival. Time Boolean Comment Column Name Graphic. File Type Comment String 16

6. 1. 4 CDR Physical Design Tsunami Forecast Earthquake Origin CMT Earthquake Origin Seis. Com. P 3 Earthquake Geometry and Size RIFT, SIFT & ATFM Tsunami Forecast Seismic data ingest And on to EDEX Seismic Processing results – Event Origin, Theta, Mw(Mm), FFS Inversion, Mwp, Array Processing, etc. Sea level data ingest 17

6. 1. 4 CDR Physical Design Seis. Com. P 3 Internal Processing SCMaster – Spread Messaging System Modules Picks Auto. Picker Teleseismic Auto. Locator Teleseismic Auto. Picker Regional Auto. Locator Regional Amplitude Magnitude Event Raw Seismic Data Real Time Data Streams Real Time Seismic Data (SEEDlink) Picks Event Origin + Amplitude + Magnitude Pick Loader = Event Quake. ML Loader Qpid Messaging Seis. Com. P 3 seismic ingest and processing easily integrates EDEX 18

6. 1. 4 CDR Physical Design CMT / RIFT / SIFT Automatic Pick Interactive Pick GUI Time Event Locator (Lat, Lon, Depth, OT) mb m. B Mwpd First Motion Focal Mech ML SIFT RIFT Tsunami Forecast Control MT_USGS MT_W-Phase 19

6. 1. 5 CDR Software Design AWIPS II and TOPS prototype reuse TOPS is built from extensive component reuse into an integrated SOA AWIPS II Reuse AWIPS II EDEX SOA Core AWIPS II CAVE Core New Code Seis. Com. P 3 TWC Reuse Storage Management Localization adapter CAVE Localization GUI Hazard services product gen TOPS Software TOPS Phase 1 Reuse AWIPS II EDEX SOA Core Seismic data adapter AWIPS II CAVE Core Seismic event (Quake. ML) Seismic filtering framework Seismic monitoring CAVE TOPS perspective CAVE bathymetry Component Reuse AWIPS Reuse Components Phase 1 Reuse components New components FOSS Packages Seis. Com. P 3 TWC Reuse New Components Sea Level Adapters TF model Adapters FOSS Reuse FOSS Packages 20

6. 1. 5 CDR Software Design Sea level ingest and processing Sea Level (SL) ingest is implemented in new EDEX plugins and leverages the ingest pipeline Nanometrics Interface Naqs. Server Process EDEX - CAVE Common Sea Level Plugin Naqs. Server Adapter SL Ingest SL Processing Reuse “c” Socket code IOC Web Service Adapter NOC Web Service Adapter TWC Local Network Spectrogram and Filtering Reuse bandpass filtering from phase 1. Spectrogram: Evaluating FFTW FOSS library with a Java wrapper. FFTW appears to be the open source standard JTransforms and apache’s Fast. Fourier. Transformer are being considered for a pure Java implementation Product Generation Tide. View / Tide. Tool Functions Re-implement SL Processing Display Tidal constants De-tiding / filtering (MATLAB) De-spiking <library> WMO Decoder Socket … Reuse “c” WMO decoder Spectrogram DBMS Tidal forecasting Detect Tsunami Auto. Pick. Tsunami Wave char. 21

6. 1. 5 CDR Software Design Sea level ingest and processing Sea Level (SL) ingest is implemented in new EDEX plugins and leverages the ingest pipeline Sea Level Input Origin and Format Comment SEEDlink servers Nanometrics Naqs. Server TSMART gauges PTWC Runup gauges Earthworm import/export WMO data formats Re-implement/reuse WMO decoder in Tide. Ingest/Src/Decoder IOC Web Service Re-implement IOC_Webservice as EDEX plugin NOS Web Service Re-implement NOS_Webservice as EDEX plugin NEPTUNE, MARS… 22

6. 1. 5 CDR Software Design Seismic Ingest and Processing Metadata is loaded from “Dataless SEED Volumes” on demand. When archived data is processed, the system meta data is replaced with historical meta data from appropriate “Dataless SEED Volumes. ” Data SEEDlink Realtime data Seis. Com. P 3 Arclink – Archived data Metadata Current Dataless Volume Seis. Com. P 3 Historical Dataless Volume 23

6. 1. 5 CDR Software Design Seismic Alarm Design and Messaging Quake. ML - Messaging Seismic Alarm Design • Quake. ML • Examples: – Events/Origins – Picks • Will Expand for: – – – CMT – W-phase – Array processing Seismic Alarms Seismic Data Products System Alarms SOH Monitoring Error Monitoring 24

6. 1. 5 CDR Software Design Messaging Seis. Comp 3 TOPS SOA Qpid Mailboxes: Earthquakes Alarms SOH Quake. ML Errors 25

6. 1. 5 CDR Software Design Tsunami forecast models General adapter pattern integrates the three tsunami models RIFT Adapter Plugin I/O Files SIFT Adapter Plugin I/O Files AFTM SL Processing SL Ingest SIFT Tsunami Forecast Views (SIFT, RIFT, AFTM) EDEX - CAVE Product Generation Display AFTM Adapter Plugin I/O Files TWC Local Network DBMS 26

6. 1. 5 CDR Software Design Control Messaging Commands Seis. Comp 3 TOPS SOA Qpid Mailboxes: configuration changes repick newpick reprocess etc. XML 27

6. 1. 5 CDR Software Design Control messaging: configuration control details Restart affected module Seis. Comp 3 Configuration & XML Configuration Module XML to Object parser Systems Monitor Update command XML to Quake. ML parser Changes Messenger Current Config TOPS SOA ● Stores configuration descriptions in XML and stores values in <Var> = <Value> format ● Single module must be restarted for new configuration to take effect 28

6. 1. 5 CDR Software Design Control messaging: command control details Seis. Comp 3 Remote Commander Module Response Messenger Remote Command Response TOPS SOA XML Translator 29

6. 1. 5 CDR Software Design Product generation Product editor edits a standard unformatted format, which generations several output formats AWIPS II Hazard Services Product Generation Google’s common event bus interface Hazard event objects for DB storage Product generation framework CAVE dialogs for product generation 30

6. 1. 5 CDR Software Design Product generation supporting multiple product formats Example XML Format Example Legacy Text Format Product generation features Unified support for multiple formats (legacy text, XML, CAP, etc. ) through tabs Interfaces to DBMS for data and GIS screen Uses a finite state machine to manage VTEC and product state 31

6. 1. 5 CDR Software Design GIS processing GIS map layers design based on ESRI shape files use AWIPS map layer approach GIS layer storage in spatially extended RDBMS enables access by all TOPS components to GIS data through spatial queries using the AWIPS core API 63 TOP ESRI shape file layers CAVE GIS display Postgis data loader shp 2 pgsql Converts ESRI into SQL for DB loading Contain > 60 Java classes of GIS operations, interfaces, APIs, and functions DBMS Postgre. SQL RDBMS with Postgis Module Postgis Spatial Tables GIS processing AWIPS common geospatial Geo. Tools and JTS FOSS libraries 32

6. 1. 5 CDR Software Design GIS processing Leveraging the GIS capability in CAVE GIS processing for layer management and map projections Display initiation, update, and capture built into CAVE Dynamic data sampling and cursor coordinate display built into CAVE GIS data interfaces to DBMS through AWIPS API Phase I addition included GPU implementation of dynamic bathymetry shading 33

6. 1. 5 CDR Software Design Historical data access Historic data seamlessly integrated into displays Historic Data Access Capability Design Approach Earthquake information stored EDEX plugin defines DBMS storage records Earthquake data search EDEX plugin defines data access objects (DAO) for queries Tsunami information stored / searched EDEX plugin defines DBMS storage records and DAOs Historic data display CAVE view for dialog box list, CAVE resource for GIS layer display Centroid Moment Tensor (CMT) display CAVE resource for GIS layer Reports CAVE view for various graphical and text data Current event auto-query Add to existing CAVE event display, add view for dialog box list 34

6. 1. 6 CDR Security Design • TOPS will meet security standards and FISMA requirements – NOAA standard images and hardening (AWIPS) – Risk based security design – Use of FISMA standards (people, process, design) • Layered design enables security in depth • Embedded use of ACLs, least privilege, monitoring, etc. – Will implement ISAs and direct SOC and N-CIRT coordination – Phase I created the secure Linux baseline so that Phase II occurs on a hardened platform. The secure baseline is stored in the Configuration Management (CM) system to ensure uniform application across the environment. TOPS Design Conforms to Latest NWS Security Requirements 35

6. 1. 7 CDR Performance / Availability Design High availability by local clusters and each site getting a superset of data OPSNet, NWWS, NWS Regional Circuits, Internet, QDDS / EIDS, GTS (through NWSTG) Ingest Cluster Each site ingests a super set of data CAVE: Tsunami Perspective Display. Processing Cluster Seismic Ingest EDEX 1 Seismic Processing EDEX 2 TWC Local Network NAS (RAID 5) DBMS HDF 5 Localization PTWC NTWC rsync localization changes between sites DB Server DB Backup Qpid Broker Postgres/ Postgis Admin GUI DB availability by heartbeat failover 36

6. 1. 7 CDR Performance/Availability Design Monitoring - State of Health Monitoring system - XML Seis. Comp 3 Heart Beat Daemon Checks to see if “Enabled Keep. Alive” modules are still running, restarts if they are not. Logs ● Changes to control modes ● Manual starts and stops ● Errors Control Utility manually start and stop Status Control TOPS SOA The following modes for control: ● Enable Keep. Alive ● Disable Status will return: ● Running ● Error ● Not Running 37

6. 1. 8 CDR Internal Communications Design Seis. Com. P 3 to EDEX interface Seis. Com. P 3 interfaces to EDEX through three specific SOA queues Seis. Com. P 3 SOA Interface Message Queues EDEX - CAVE Quake. ML Plugin CAP Alarm Plugin Seis. Com. P 3 Controls Interface SL Ingest SL Processing Quake. ML Generate CAP Alarm Generate General Input Interface Product Generation Display TWC Local Network DBMS Seis. Com. P 3 to EDEX Interfaces Description Quake. ML to EDEX Initially implemented Phase I (com. raytheon. uf. edex. plugin. quakeml. ingest). Message queue with the Quake. ML schema defining the payload. Communicates earthquake events details. Will add CMT results. CAP Alarm to EDEX Message queue with the CAP XML schema containing details of seismic events that need operator attention. EDEX to Seis. Com. P 3 Message queue containing a generate purpose payload of name/value pairs. Used to transfer operator re-picks and seismic configuration changes. 38

6. 1. 8 CDR Internal Communications Design Internal Interfaces using Quake. ML Any Seismic Detection System Seis. Com. P 3 Any Seismic Detection System Quake ML Seismic Interface XML Pick Loader Event Loader Repick Messages Commands/ relocating Qpid Messaging Seis. Com. P 3 TOPS SOA Qpid Messaging Seismic detection modularity and the use of seismic community standards TOPS SOA Quake. ML enables easy integration with Seis. Com. P 3 and other leading seismic systems 39

6. 1. 8 CDR Internal Communications Design Seis. Com. P 3 as a library for TOPS SOA • Filtering - Allow for TOPS SOA GUI to directly use proven key Seis. Com. P 3 functionality to TOPS SOA via API. • Ensures that filters used by the GUI are functionally identical to filters used in Seis. Com. P 3 processing. 40

6. 1. 8 CDR Internal Communications Design Seis. Com. P 3 Signals Processing Seis. Comp 3 Detection Windowing STA/LTA Cosine Taper Filtering Butterworth Filters: • BW Lowpass • BW Highpass • BW Bandpass • BW Notch Other Filters: • Detrend • Integration • Differentiation • Running Average Simulation Wood Anderson WW SSN Short Period WW SSN Long Period 41

6. 1. 8 CDR Internal Communications Design Tsunami forecast RIFT interface Integrate RIFT through an EDEX adapter, minimizing RIFT changes rift. exe rift_gui. py RIFT Adapter Plugin EDEX - CAVE SL Processing SL Ingest Product Generation Display Rift Interface Files Source. in Rift. in Tide. out Basin. in Warn. out Cmt params Coast. out Diag. out Rift. out Run summary Job. in Source. out Job. out topo TWC Local Network DBMS 42

6. 1. 9 CDR System Data Integrity Controls • • Turning on kernel-level auditing to capture individual user actions on the system in log files; these actions include successful/unsuccessful logins and logoffs, unauthorized file access attempts, escalation of user privileges, user account creation/deletion/modification, service startup/shutdown, and other significant events. Users can only login to the system with individual user accounts and passwords that meet vetted password strength requirements. Escalated privileges are granted only if the user has been authorized for escalated privileges (e. g. , root access). Users without privileged access only have read access to certain files, and have limited access to applications. Users who need escalated privileges will not have those rights assigned to their standard user accounts, but will “SU” to the privileged account only for the duration of the work requiring such rights. Remote access to the system is only via secure protocols. No user credentials are stored in clear text. Files are configured not to be openly writable unless they are needed for the system mission. Enabling the integrity-checking software (AIDE) in the operating system to notify system administrators of any unauthorized modification of critical or security-relevant files. 43

6. 1. 10 CDR User Interface Design Plan Consolidated display design approach PDR concept of integrating TOPS displays into CAVE perspective CAVE: Tsunami Perspective Real-time Seismic Geographic (2) Seismic Analysis Received Products Real-time Sea level Event Management Associated event data Forecasting display System Status Display GUI Configuration and Localization Common Site-PTWC Site-ATWC Thin Client User. . . Interactive Event control 44

6. 1. 10 CDR User Interface Design Plan Extending AWIPS CAVE core to TOPS displays CAVE core (12 AWIPS reuse plugins) Relocatable Views Click on Event / Alarm triggered Views are all event driven and continually update as messages come in Event Locator View Event Details View Display maintains state through shutdowns Magnitude Calculator View GIS View Seismic Timeseries and Picker View 45

6. 1. 10 CDR User Interface Design Plan Event Details View contains buttons for tsunami forecast and product generation views Location magnitude type time section RIFT- Forecast Button… Product Generation Button Past event table section RIFT Views SIFT Views Product Generation Views AFTM Views 46

6. 1. 10 CDR User Interface Design Plan TOPS display screen shots TOPS display views consolidated into CAVE Plugins Current Defined Plugins Planned CAVE core (12 AWIPS reuse plugins) CAVE tsunami forecast CAVE seismic (time series view) CAVE Product generation CAVE seismic event CAVE seismic location CAVE Alert display CAVE seismic magnitude CAVE seismic picker (like PTWC’s) CAVE topo bathymetry CAVE travel times CAVE tsunami perspective 47

6. 1. 11 CDR External Interfaces PDR external interfaces update Real Time Seismic Data Streams Real Time Seismic Data (SEEDlink) Outputs Seismic Processing (Seis. Com. P 3 ) Picks Pick Loader Events Quake. ML Loader SIFT RIFT TOPS SOA Sea Level Processing Products Warning Products Tsunami Forecast ATFM DBMS Qpid Messaging Operator Input Web Real Time Sea Level DART Buoys Display (CAVE) Product Generation 48

6. 1. 11 CDR External Interfaces Interface implementation design concept TOPS is built entirely from components using a single plugin pattern maximizing reuse and reliability Data Adapters SOA Core FOSS Single Standard Plugin Design Pattern Plug-in Implements Ingest Adapter Plugin Data Decoder Metadata Record Processing Services Ingest Pipeline Services Data Access Objects CSS-Wx Internal Service Bus (Extendable) Data Archive Algorithm / Transforms Hosted Algorithm Plugin Scripts Libraries Storage Rules/ Purge Plugin Pattern Interfaces Plugin Pattern • • • Algorithms Transformation Services Request Services SOA Core C++ Libary Product Generation Storage Management Services/ APIs Output Adapter Plugin DBMS TOPS Storage Product Gen Data/Interface Adapters TOPS software is created by adding new plugins to the reuse plugins Plugin pattern handles all component types (FOSS, SOA infrastructure, data adapters, algorithms) Plugin component built with a standard (Eclipse IDE) Software Development Kit (SDK) SDK kit packages the plug-in into a Java JAR file with managed versioning SDK enables re-use of core design patterns and infrastructure 49

6. 1. 11 CDR External Interfaces Top level interfaces External interface adapters Seis. Com. P 3 Seismic Ingest EDEX TOPS Server SEEDlink Server Naqs. Server SEEDlink Adapter Plugin Nanometrics Adapter Plugin Seismic Adapter Plugin Sea Level Adapter Plugin TSMART Gauges WMO Formats PTWC Runup Import / Export Web Service NEPTUNE/MARS PTWC Runup Adapter Plugin Earthworm Adapter Plugin Product Generation WMO Adapter Plugin Ingest Pipeline Services IOC /NOC Web Adapter Plugin Web GUI Framework Request Services OGC Services Consumer Adapter Plugin NEPTUNE. . Adapter Plugin DBMS Storage Management Services/ APIs TOPS Internal Service Bus (Extendable) Transformation Services TSMART Adapter Plugin Realtime Metadata (postgre. SQL) Data Config Data Scripts (Files) 50

6. 1. 11 CDR External Interfaces Sea-level input interfaces Design / Implementation Approach Legacy Source PTWC Tide. tcl Re-implement WMO text decoding Sea level data ingest adapter plugins to EDEX WMO Formats WMO Adapter Plugin Sea level adapter integrated into EDEX Naqs. Server Legacy Source Wcatwc_Earthworm Re-implement Naqs. Ser. TG Nanometrics Adapter Plugin Web Service Legacy Source NTWC Tide. Ingest Re-implement IOC/NOC Webservice IOC /NOC Web Adapter Plugin Earthworm Import / Export ISTI Earthworm Source Adapter Re-use Legacy Source PTWC neptune Re-implement get_neptune_dat. tcl Sea Level Adapter Plugin Ingest Pipeline Services Storage Management TOPS Internal Service Bus (Extendable) Services/ APIs DBMS Realtime Metadata (postgre. SQL) Earthworm Adapter Plugin Data NEPTUNE/MARS Config Data Scripts (Files) NEPTUNE. . Adapter Plugin 51

6. 1. 11 CDR External Interfaces Seismic and Sea level output interfaces Sea level output interface Sea level output adapter to NGDC Seismic Processing export to IRIS DMC • • • SDP will export data to other seismic data repositories including: – NEIC – IRIS DMC – University of Alaska, Fairbanks – Alaska Volcano Observatory – University of Hawaii – Hawaii Volcano Observatory This export will be via SEEDlink or Quake. ML or Import/Export. Sea Level may also be exported via an output adapter to NGDC. 52

6. 1. 11 CDR External Interfaces Product generation external interfaces Product Generation Dissemination Interface Design Approach GTS via NWS Telecom EMWIN via NWS Telecom RANET via NWS Telecom AFTN (NADIN 2) via NWS Telecom NOAA Weather Wire OPSNet NWS lines Internet to web sites EDEX apache. Camel routes RSS feeds EDEX apache. Camel routes EIDS Earthquake Information Distribution System Email list + FAX EDEX apache. Camel routes SMS message lists EDEX apache. Camel routes TWC phone system Reuse existing phone interface IPAWS / EAS via NWS telecom messages CAP format (profile) 53

6. 1. 11 CDR External Interfaces GIS interfaces GIS Interfaces Design Approach NGDC Historic Tsunami database DBMS EDEX API NEIC Historic earthquake database DBMS EDEX API GVP (Smithsonian) Volcano database DBMS EDEX API 54

6. 1. 11 CDR External Interfaces Historical data interfaces All historical data is ingested into the DBMS through adapters Historical data interface Design Approach USGS-NEIC for > 3. 0 Mag EDEX plugin interface adapter Smithsonian Institution Global Volcanism Program EDEX plugin interface adapter CMT database EDEX plugin interface adapter National Geophysical Data Center tsunami database (NGDC – NOAA) EDEX plugin interface adapter Update quarterly 55

6. 1. 12 CDR Data Migration / Conversion Plan • Tools were written/modified in Phase I to convert from local data formats to miniseed format. • Miniseed is the data-only portion of a SEED Volume. • SEED - Standard for the Exchange of Earthquake Data. • Historic data will be stored in SEED format and served to Seis. Com. P 3 via arclink servers. • Additional converters will be written, as needed. RAW Data Create Tracebuf Data NTWC Data ATPlayer Create miniseed files tank 2 mseed PTWC Data Miniseed CSS 2 Tank 56

6. 1. 13 CDR Implementation / Release Plan • TOPS deployment, installation, and implementation will follow our Project Plan (Transition Plan) and approved project schedule (Work Breakdown Structure). • Parallel implementation and phased transition will foster confidence before going live. • Coordination and cooperation between the ERT Team, NWS and TWCs; user acceptance; demos; site visits; and training activities will support a successful implementation. 57

6. 1. 14 CDR Infrastructure Version Tracking Plan • TOPS baseline updates will be tracked and integrated into the final, deployed system. • Baseline updates will follow our Project Plan (Configuration Management [CM] Plan, Requirements Management Plan, and Software Development Plan). • We will run scans regularly to check for any changes in our hardware and software baseline using CM tools. • If a change is detected, it will be tracked in Redmine and reviewed by the Change Control Board (CCB). • At the time of deployment into operations, TOPS will use supported versions of hardware and software infrastructure components. 58

6. 1. 15 CDR Hardware Configuration Recommendation Reduce TOPS Cost by Reusing WFO Hardware for TOPS Software Components LXn CAVE TWC Local Network (bus) Hot Failover Clustering Load Balance Clustering Partition Clustering DX 1 DX 2 DX 3 DX 4 PX 1 PX 2 Postgre. SQL EDEX-Ingest Seis. Com. P 3 QPID EDEX-Request Quake. ML Adapter Py. Pies DAS DBMS 59

WRAP UP 60

Wrap Up • Questions? • Consolidate action items • Comments and corrections 61