230 likes | 345 Views
HarmoniRiB Workshop on Uncertainty in Data and Models Brussels, September 21, 2006 Design of a Water Information System for future needs Thomas Bech (DHI Water & Environment) Roger V. Moore (CEH) Giuseppe Passarella (IRSA). Outline. Trends and Future Needs The HarmoniRiB database Conclusions.
E N D
HarmoniRiB Workshop on Uncertainty in Data and Models Brussels, September 21, 2006Design of a Water Information System for future needsThomas Bech (DHI Water & Environment)Roger V. Moore (CEH)Giuseppe Passarella (IRSA) HarmoniRiB Workshop on Uncertainty, Brussels, September 21, 2006
Outline • Trends and Future Needs • The HarmoniRiB database • Conclusions HarmoniRiB Workshop on Uncertainty, Brussels, September 21, 2006
Trends in Information and Decision Support Systems HarmoniRiB Workshop on Uncertainty, Brussels, September 21, 2006
Recap - Limitations of current databases • No possibility to store information on data uncertainty • Mostly national datasets • different terminology (dictionaries) across countries • different variables and measurement techniques across countries • Mostly single domain data • need for all domain data (meteorology, hydrology, geology, ecology, socio-economics, etc) in one database allowing easy cross-sectoral analysis HarmoniRiB Workshop on Uncertainty, Brussels, September 21, 2006
Trends • We are moving from dealing with • Single, isolated problems • To • Complex, multi-disciplinary issues • Symptoms • Data -> Information -> Knowledge -> Decision • Reports -> Operational Systems • Models -> Information Systems -> Decision Support Systems HarmoniRiB Workshop on Uncertainty, Brussels, September 21, 2006
Example: Hydropower DSS The user: Operates the hydropower facility • The user want a system that: • Provides the predicted revenue from power sales for the next 48 hours, with a given release strategy. • Provides the predicted revenue from power sales for the next 48 hours, with a altered release strategy, power prices and precipitation rate (what if scenarios) HarmoniRiB Workshop on Uncertainty, Brussels, September 21, 2006
Example: Hydropower HarmoniRiB Workshop on Uncertainty, Brussels, September 21, 2006
DSS Key Characteristics • Data integration across domains, types and sources • Combines, aggregates, analyses data • Presents information useful for decision making • Keywords: Integration – Analysis – Presentation HarmoniRiB Workshop on Uncertainty, Brussels, September 21, 2006
Future Needs - Summary • We need frameworks that allows us to link “DSS Components” in a nearly seamless manner (such as OpenMI) • We need data integration platforms which are adaptable, scalable and Open • We need a portfolio of “Plug-and-play DSS components” • Uncertainty must be addressed – in the data integration platform as well as the tools HarmoniRiB Workshop on Uncertainty, Brussels, September 21, 2006
The HarmoniRiB Database HarmoniRiB Workshop on Uncertainty, Brussels, September 21, 2006
The HarmoniRiB database Rep Basin managers Europe Rep Basin managers Data Centre Search Request Retrieve Rep Basin managers Suppliers Rep Basin managers Users Rep Basin managers Office based Geographically dispersed suppliers and users HarmoniRiB Workshop on Uncertainty, Brussels, September 21, 2006
The HarmoniRiB database • Design is based on CEH’s extensive experience in Water Information Management Systems and CEH’s existing systems • Is implemented using proven technologies and software (ORACLE, ArcSDE/ArcGIS) • Implements a generic approach to data storage • Uncertainty information added using the DUE • Is operational at the project data centre (IRSA at Bari, Italy) • Project data from river basins across Europe available and can be retrieved through Web interface or SQL HarmoniRiB Workshop on Uncertainty, Brussels, September 21, 2006
Software Developed • The database • Database administration software • Data Loader Software • Web site for Uploading / Loading data • Web site for data retrieval HarmoniRiB Workshop on Uncertainty, Brussels, September 21, 2006
Object Attribute A generic approach to storage • All data stored as attributes of objects • All data are assumed to change over time • No distinction made between spatial and non-spatial attributes • Dictionaries of attributes and objects can be extended. i.e. Scalable HarmoniRiB Workshop on Uncertainty, Brussels, September 21, 2006
Attributes and attribute dictionaries • Attributes are used to record “what” has been described or observed at an object • Examples of attributes are: • Object class • Site ID • Site name • Location (X,Y, Z) • Width • Water level • Concentration of mercury • Fish count • Mean daily river flow • Species ……. abundance HarmoniRiB Workshop on Uncertainty, Brussels, September 21, 2006
Attributes and attribute dictionaries • Attributes can be numerous – e.g. species lists or chemicals • Attribute definitions are therefore grouped in to dictionaries • Example dictionaries are: • Chemical • Hydrological • Hydrogeological • Algal • Macrophytes • Invertebrates • Weather • Universal • System HarmoniRiB Workshop on Uncertainty, Brussels, September 21, 2006
The Data Centre StatusDatasets • Number of records loaded per Dictionary HarmoniRiB Workshop on Uncertainty, Brussels, September 21, 2006
The Data Centre StatusDatasets • Number of records loaded per River Basin HarmoniRiB Workshop on Uncertainty, Brussels, September 21, 2006
1.0 0 Quantitative modelUncertainty recorded as a PDF • PDF = Normal • Mean = • SD = σ ? ? HarmoniRiB Workshop on Uncertainty, Brussels, September 21, 2006
Conclusions Confirmed: • The HarmoniRiB database has proven sustainable as the data platform of the HarmoniRiB project (including WFD data) • The HarmoniRiB database platform is a suitable data integration platform for future DSS / Water Information Management systems Challenges: • The database is not suitable for low-end systems • To operate it successfully requires IT-proficient staff – and thorough understanding of the data model HarmoniRiB Workshop on Uncertainty, Brussels, September 21, 2006
Future Needs - recap • We need frameworks that allows us to link “DSS Components” in a nearly seamless manner (such as OpenMI) • We need data integration platforms which are adaptable, scalable and Open • We need a portfolio of “Plug-and-play DSS components” • Uncertainty must be addressed – in the data integration platform as well as the tools HarmoniRiB Workshop on Uncertainty, Brussels, September 21, 2006
Next steps - exploitation • Transformation to operational product • Improved GUI • Improved data loading facilities • Keeping the database alive • Hosting • Maintenance • Initial set of applications by building collaborations with e.g. • LOCAR (UK) • AQUASTRESS (EU) • CUAHSI (US) HarmoniRiB Workshop on Uncertainty, Brussels, September 21, 2006
Messages • Integration of data across types / domains is a component of every Decision Support System • Uncertainty must be addressed • The database developed in HarmoniRiB deals with both HarmoniRiB Workshop on Uncertainty, Brussels, September 21, 2006