Jane Hunter,
The University of Qld
Microsoft eScience 2008

 Background and Objectives
 Architecture and Implementation
 User interface and functionality
 Demo
 Conclusions and Future Work
Microsoft eScience 2008

 Collaboration between:
 Microsoft Research (Catharine van Ingen)
 Healthy Waterways Partnership (Eva Abal)
 DNRW, EPA, Local Councils, Universities
 University of Qld (Jane Hunter)
 3 years funding – MSR, ARC Linkage, SmartState
 Integrated Water Information Management for
SEQ-HWP

 Fast growing population
 Severe water shortages
 Sensitive ecosystems
 Climate change and drought

127 freshwater sites (sampled 2x/yr)
254 estuarine and marine sites
(sampled monthly)

 FreshWater EHMP - Dept. Natural Resources and
Water (DNRW)
 Estuarine Marine EHMP - EPA
 Event Monitoring – DNRW
 Management Action Database – SEQ-HWP
 Models – many different sources/locations
 Receiving Water, EMSS, E2




The data is being captured and managed by DNRW
127 freshwater sites across the catchments.


16 Indicators from 5 categories:


Physical and chemical – pH, Conductivity, temp, dissolved O
2
Nutrients Ratio of nitrogen stable isotopes ( δ 15 N), algal growth


Ecosystem processes Algal growth, Ratio of carbon stable isotopes ( δ 13 C), Benthic respiration (R
24
) Primary production GPP
Aquatic macroinvertebrates – No. taxa, PET, SIGNAL
 Fish% of native species expected (PONSE), Observed to expected native species (O/E
50
), Proportion of alien fish
Surveys are conducted every 6 months, spring and autumn.
Survey data stored in Oracle relational database.






The data is being captured and managed by the Environmental
Protection Agency
254 Sites in South East Queensland:


168 sites from 19 estuaries
86 from Moreton Bay
14 Indicators :




Turbidity , Salinity, Temperature, Dissolved Oxygen, pH, Secchi depth, Nitrogen, Phosphorus and Chlorophyll.
Lyngbya Majuscula (seaweed) cover.
Sewage plume mapping
Coral Cover
Surveys are conducted monthly, biannually and annually.
Survey results will be stored in an Oracle relational database.







The data is being captured and managed by the Dept of
NRW
60 to 100 sites across South East Queensland
Proprietary software known has HYDSTRA by the Kisters group is used to store the data
Compressed files store time-series data for each site
 River height, Daily Min/Mean/Max flow


Pollutants
Events - floods
Supporting information is also stored:
 E.g. water parameters, survey technicians
Raw data is less useful than interpreted data

 Managed by SEQ-HWP
 Tracks Action Plans that are part of the Healthy
Waterways Strategy
 Approximately 550 actions are stored in the database
 2003 Access database:
 Access relational tables back-end
 Access forms front-end
 Interface and actions is organised through a 4 tier hierarchy

 Many different models used for catchment hydrology
 The model simulations forecast and emulate climate scenarios
 Written in many different languages for a variety of purposes and users - Fortran
 Focus on 3 Models:
 EMSS (Environmental Management Support System)
Catchment Model
 Receiving Water Model
 E2

EHMP Estuarine/Marine
Remote
Sensors
EHMP Freshwater
EHMP Event Monitoring
Model scenarios, outputs
Management Actions
SEQ Water
Bureau of
Meteorology
Landuse
Demography
Etc.
Example Query:
What will be the ecosystem health outcomes of the implementation of landscape restoration works in the
Logan Albert System by 2026?
Health-e-Waterways
• Web Portal
•
Water Wiki
•
VirtualEarth
• SensorMap
R
I
T
Y
S
E
C
U
•
Data Ontology and Server
• Web Services
• Data Integration
• Data Lineage
• Uncertainty Propagation
• Models and Workflows
L
A
Y
E
R
General Public
State Government
Local Governments
Water Resource
Managers
Researchers
Scientists
Hydrologists
QCIF Grid Computing &Storage www.healthywaterways.org



Streamline Annual EHMP Report Card Generation
Search, analysis, reporting interface to integrated databases


 Identify common conceptual model (ODM, OpenGIS,
WRON-RM)
Map datasets to common model
Identify optimum data harvesting and storage
 Store in SQLServer DB or Jena
 Web services interface to in-situ data
 Metadata harvesting -> central catalogue
 Develop VirtualEarth+ontology-based query interface

 Publicised output of the
SEQ Healthy Waterways
Partnership
 Easy-to-understand snapshot of ecosystem health
 A to F
 Provides an insight into the effectiveness of investments in waterway and catchment management
 Split into two reporting zones, freshwater and estuarine/marine
 Each has it’s own objectives, parameters,
What methods and analysis

FRESHWATER REPORT CARD GRADES
Pumicestone Catchment
Grade history:
# combined grade for Caboolture-Pumicestone catchments
2001 2002 2003 2004 2005 2006 2007
C
#
C
#
B+
#
C C+ CC-
Spring 2005 (Run 7) Autumn 2006 (Run 8)
 snapshot of ecosystem health

A to F
 insight into the effectiveness of investments in catchment management
Spring 2006 (Run 9) Autumn 2007 (Run 10)

How

EHMP Ontology
Interactive Ecosystem Report Card Application
Reasoning
Client
Silverlight &
Virtual Earth Client
SPARQL
Query Client
Reasoning Engine
Web Services
Statistical Processing
Triple Store Jena .NET Plugin
Remote Sensor
Administrator
EHMP Databases

 Common Observational Data Model


ODM 2.0?
Water data, climate data, vegetation, species distribution, satellite imagery
 Framework for Semantic Integration of Ecosystem Health
Monitoring Data
 ICT Framework for Web-based Environmental Reporting
 Standardized methods for measuring and aggregating indicators -> Ecosystem reports

 Comparison and longitudinal trends
Wentworth Group – “an exemplar for environmental reporting”








Link monitoring data to management actions
Integration of:
•
•
MODIS satellite data, BoM climate data
Real-time sensor data
•
•
Community data – ReefCheck, CoralWatch, Caring for Country
Socio-economic data - demographics
Extend to Great Barrier Reef /Centre for Marine Studies
Analytical services
 correlate ground data to derived data from satellite images
Linking predictive models to integrated datasets
Visualizations of model output
Estimate uncertainty/reliability of results

Ranked search results

Combine monitoring data + Model outputs + socioeconomic models/data
 “How will the mandatory adoption of rainwater tanks in the Logan Region effect domestic water requirements in
5 years time, taking into account the effects of climate change and population growth in the region?”
 “What impact will a $20mill sewage treatment plant upgrade have on on the prawn industry in the Logan
Estuary if implemented now?”






Abdul Alabri – University of Qld
Microsoft Research – Catharine van Ingen, Bora Beran
Healthy Waterways Partnership – Eva Abal, Jo Burton, Dave
Moffat
CUAHSI – Dave Maidment, Michael Piasecki
CSIRO – Simon Cox, AWRIS

http://www.health-e-waterways.org/
Contact: j.hunter@uq.edu.au