H13H-1679 

Construction of Evolutionary Artificial Neural Networks and its Application to Streamflow Forecasting

Chang, F (changfj@ntu.edu.tw), National Taiwan University, No.1,Sec.4,Roosevelt Road, Taiwan(R.O.C.), Taipei, 10617, * Chen, Y (wesleychen@wra.gov.tw), National Taiwan University, No.1,Sec.4,Roosevelt Road, Taiwan(R.O.C.), Taipei, 10617,

Selection of architecture for an artificial neural network (ANN) has significant influence on its successful application to the tasks to be performed. However, the conventional way for an ANN to process the optimization search is first to predefine fixed architecture and learning rule and then experience a series of trials and errors. This way usually leads the search to a local optimum and lack of efficiency and robustness. To improve the drawbacks of the conventional optimal process, this study introduces a novel algorithm, evolving artificial neural networks. With a hybrid encoding of network architecture, we apply genetic algorithm to optimize the encoded parameters of feedforward ANNs' architecture. Then we optimize the connection weights of neurons by the scaled conjugate gradient algorithm. The terrific performance for forecasting of Mackey-Glass chaotic time series shows that the proposed algorithm concurrently possesses efficiency, effectiveness, and robustness. Besides, the application to the forecasting of 10-day reservoir inflows reveals again the algorithm's excellent efficiency and robustness and its effectiveness which is superior to that of the AR(1) and ARMAX models. Keywords: Evolutionary artificial neural network (EANN), Genetic algorithm (GA), Hydrological systems, Forecasting, Reservoir inflow

H13H-1680 

Development of GIS-Based Database and Vulnerability Analysis Tools for Mississippi Groundwater, Surface Water, and Dams

* Holtz, T (tholtz@intera.com), INTERA, Inc., 1812 Centre Creek Dr., Suite 300, Austin, TX 78754, United States Pickens, J (jpickens@intera.com), INTERA, Inc., 1812 Centre Creek Dr., Suite 300, Austin, TX 78754, United States Jones, T (tjones@intera.com), INTERA, Inc., 1812 Centre Creek Dr., Suite 300, Austin, TX 78754, United States Holt, R M (rmholt@olemiss.edu), Department of Geology and Geological Engineering, University of Mississippi, 118 Carrier Hall, University, MS 38677, United States Kuszmaul, J (jkuszmaul@olemiss.edu), Department of Geology and Geological Engineering, University of Mississippi, 118 Carrier Hall, University, MS 38677, United States Gunter, B (bjgunter@olemiss.edu), Department of Geology and Geological Engineering, University of Mississippi, 118 Carrier Hall, University, MS 38677, United States Phillips, P (pat_phillips@deq.state.ms.us), Office of Land and Water Resources, Mississippi Department of Environmental Quality, PO Box 10631, Jackson, MS 39289-0631, United States Cobb, D (david_cobb@deq.state.ms.us), Office of Land and Water Resources, Mississippi Department of Environmental Quality, PO Box 10631, Jackson, MS 39289-0631, United States Caviness, D (don_caviness@deq.state.ms.us), Office of Land and Water Resources, Mississippi Department of Environmental Quality, PO Box 10631, Jackson, MS 39289-0631, United States Caviness, D (don_caviness@deq.state.ms.us), Mississippi State Department of Health, 570 East Woodrow Wilson Drive, Jackson, MS 39216, United States Walters, K (Karen.Walters@msdh.state.ms.us), Mississippi State Department of Health, 570 East Woodrow Wilson Drive, Jackson, MS 39216, United States

The development of a GIS-based database, or geodatabase, for the purposes of creating a state-wide inventory of groundwater, surface water and dam resources in Mississippi has been undertaken. The database supports the development of vulnerability assessments for these resources based on their susceptibility to natural and anthropogenic threats. This investigation requires the integration of a variety of datasets (e.g., elevation, hydrology, dams, geology, soils, population, transportation, etc.), made available by federal and state government agencies, into a SQL Server enterprise geodatabase. The project data model leverages from existing ArcGIS data models, including ArcHydro and NHDinGeo, for the development of a scale-independent geodatabase to support vulnerability assessment tools. ArcGIS ModelBuilder is used in conjunction with custom designed Graphical User Interfaces (GUIs) in the construction of an ArcGIS Desktop toolbox. The custom tools automate the development and manipulation of model inputs, evaluation of outputs, maintenance of the project geodatabase, the development and evaluation of models and model scenarios, and the implementation of vulnerability assessments. The geodatabase data model and tools are flexible and scaleable, allowing for the user to update, define and refine models and model input parameters.

H13H-1681 

Dynamic Multicriteria Evaluation of Conceptual Hydrological Models

* de Vos, N J (n.j.devos@tudelft.nl), Water Resources Section, Delft University of Technology, Stevinweg 1, Delft, 2600GA, Netherlands Rientjes, T H (rientjes@itc.nl), Department of Water Resources, International Institute for Geo-Information Science and Earth Observation, Hengelosestraat 99, Enschede, 7500AA, Netherlands Fenicia, F (fenicia@lippmann.lu), Water Resources Section, Delft University of Technology, Stevinweg 1, Delft, 2600GA, Netherlands Fenicia, F (fenicia@lippmann.lu), Public Research Center – Gabriel Lippmann, 41 Rue du Brill, Belvaux, L-4422, Luxembourg Gupta, H V (hoshin.gupta@hwr.arizona.edu), Department of Hydrology and Water Resources, University of Arizona, 1133 E James E. Rogers Way, Tucson, AZ AZ85721, United States

Accurate and precise forecasts of river streamflows are crucial for successful management of water resources and under the threat of hydrological extremes such as floods and droughts. Conceptual rainfall–runoff models are the most popular approach in flood forecasting. However, the calibration and evaluation of such models is often oversimplified by the use of performance statistics that largely ignore the dynamic character of a watershed system. This research aims to find novel ways of model evaluation by identifying periods of hydrologic similarity and customizing evaluation within each period using multiple criteria. A dynamic approach to hydrologic model identification, calibration and testing can be realized by applying clustering algorithms (e.g., Self-Organizing Map, Fuzzy C-means algorithm) to hydrological data. These algorithms are able to identify clusters in the data that represent periods of hydrological similarity. In this way, dynamic catchment system behavior can be simplified within the clusters that are identified. Although clustering requires a number of subjective choices, new insights into the hydrological functioning of a catchment can be obtained. Finally, separate model multi-criteria calibration and evaluation is performed for each of the clusters. Such a model evaluation procedure shows to be reliable and gives much-needed feedback on exactly where certain model structures fail. Several clustering algorithms were tested on two data sets of meso-scale and large-scale catchments. The results show that the clustering algorithms define categories that reflect hydrological process understanding: dry/wet seasons, rising/falling hydrograph limbs, precipitation-driven/ non-driven periods, etc. The results of various clustering algorithms are compared and validated using expert knowledge. Calibration results on a conceptual hydrological model show that the common practice of single-criteria calibration over the complete time series fails to perform adequately in all periods or on all criteria. Subsequently, improved model structures are constructed and the evaluation repeated. We conclude that a dynamic, multi-criteria approach to model identifying and testing is effective in constructing models that are more accurate and precise in forecasting streamflow.

H13H-1682 

Linking hydrologic models and data: The OpenMI approach

* Goodall, J L (goodall@sc.edu), University of South Carolina, Department of Civil and Environmental Engineering, 300 Main Street, Columbia, SC 29208, United States Robinson, B F (bella.robinson@csiro.au), CSIRO, Information and Communication Technologies, GPO Box 664, Computer Science and Information Technology Building, Australian National University, North Road Acton, Canberra, ACT 2601, Australia Shatnawi, F M (fmshatna@engr.sc.edu), University of South Carolina, Department of Civil and Environmental Engineering, 300 Main Street, Columbia, SC 29208, United States Castronova, A M (castrona@engr.sc.edu), University of South Carolina, Department of Civil and Environmental Engineering, 300 Main Street, Columbia, SC 29208, United States

Modeling frameworks provide the ability to create open, flexible modeling systems where simulations can be constructed from a set of computational modules interlinked for a given application. Various modeling frameworks have been proposed and developed with varying degrees of success. This paper investigates a more recent modeling framework, the Open Modeling Interface (OpenMI). OpenMI, which is freely available and open source (http://www.openmi.org), was developed by a consortium of public and private organizations in Europe. Various models are either already or in the process of becoming OpenMI compliant (e.g. MIKE-SHE, HEC-RAS, and Modflow). While using OpenMI to link existing models is one application of the framework, we are interested in determining the appropriateness of OpenMI as the underlying architecture for a community based hydrologic modeling system. We investigate this question by showing how the framework would allow a simulation model and database to be coupled, how OpenMI orchestrates the communication between these two components, and how OpenMI could be expanded to accommodate models and databases that are exposed as web services. We conclude with a discussion of the advantages and disadvantages of the OpenMI modeling framework and provide a vision for how a community model might be structured using an OpenMI-based approach.

H13H-1683 

Estimation of Missing Precipitation Data using Soft Computing based Spatial Interpolation Techniques

* Teegavarapu, R S (ramesh@civil.fau.edu), Florida Atlantic University, 777 Glades Road, Bldg # 36, Room 217, Boca Raton, FL 33431, United States

Deterministic and stochastic weighting methods are the most frequently used methods for estimating missing rainfall values at a gage based on values recorded at all other available recording gages. Traditional spatial interpolation techniques can be integrated with soft computing techniques to improve the estimation of missing precipitation data. Association rule mining based spatial interpolation approach, universal function approximation based kriging, optimal function approximation and clustering methods are developed and investigated in the current study to estimate missing precipitation values at a gaging station. Historical daily precipitation data obtained from 15 rain gauging stations from a temperate climatic region, Kentucky, USA, are used to test this approach and derive conclusions about efficacy of these methods in estimating missing precipitation data. Results suggest that the use of soft computing techniques in conjunction with a spatial interpolation technique can improve the precipitation estimates and help to address few limitations of traditional spatial interpolation techniques.

H13H-1684 

OLAP Cube Visualization of Hydrologic Data Catalogs

Zaslavsky, I (zaslavsk@sdsc.edu), San Diego Supercomputer Center, University of California, San Diego, La Jolla, CA 92093- 0505, United States Rodriguez, M (MKRodriguez@lbl.gov), San Diego Supercomputer Center, University of California, San Diego, La Jolla, CA 92093- 0505, United States Beran, B (bora.beran@gmail.com), Microsoft Research, 455 Market St., Suite 1690, San Francisco, CA 94105, United States Valentine, D (valentin@sdsc.edu), San Diego Supercomputer Center, University of California, San Diego, La Jolla, CA 92093- 0505, United States * van Ingen, C (vaningen@windows.microsoft.com), Microsoft Research, 455 Market St., Suite 1690, San Francisco, CA 94105, United States Wallis, J C (jwallisi@ucla.edu), Center for Embedded Networked Sensing, University of California, Los Angeles, Los Angeles, CA 90095-1596, United States

As part of the CUAHSI Hydrologic Information System project, we assemble comprehensive observations data catalogs that support CUAHSI data discovery services (WaterOneFlow services) and online mapping interfaces (e.g. the Data Access System for Hydrology, DASH). These catalogs describe several nation-wide data repositories that are important for hydrologists, including USGS NWIS and EPA STORET data collections. The catalogs contain a wealth of information reflecting the entire history and geography of hydrologic observations in the US. Managing such catalogs requires high performance analysis and visualization technologies. OLAP (Online Analytical Processing) cube, often called data cubes, is an approach to organizing and querying large multi-dimensional data collections. We have applied the OLAP techniques, as implemented in Microsoft SQL Server 2005, to the analysis of the catalogs from several agencies. In this initial report, we focus on the OLAP technology as applied to catalogs, and preliminary results of the analysis. Specifically, we describe the challenges of generating OLAP cube dimensions, and defining aggregations and views for data catalogs as opposed to observations data themselves. The initial results are related to hydrologic data availability from the observations data catalogs. The results reflect geography and history of available data totals from USGS NWIS and EPA STORET repositories, and spatial and temporal dynamics of available measurements for several key nutrient-related parameters.

H13H-1685 

Data Access System for Hydrology

* Whitenack, T (whitenac@sdsc.edu), San Diego Supercomputer Center, University of California, San Diego, La Jolla, CA 92093- 0505, United States Zaslavsky, I (zaslavsk@sdsc.edu), San Diego Supercomputer Center, University of California, San Diego, La Jolla, CA 92093- 0505, United States Valentine, D (valentin@sdsc.edu), San Diego Supercomputer Center, University of California, San Diego, La Jolla, CA 92093- 0505, United States Djokic, D (ddjokic@esri.com), ESRI, 380 New York St, Redlands, CA 92373, United States

As part of the CUAHSI HIS (Consortium of Universities for the Advancement of Hydrologic Science, Inc., Hydrologic Information System), the CUAHSI HIS team has developed Data Access System for Hydrology or DASH. DASH is based on commercial off the shelf technology, which has been developed in conjunction with a commercial partner, ESRI. DASH is a web-based user interface, developed in ASP.NET developed using ESRI ArcGIS Server 9.2 that represents a mapping, querying and data retrieval interface over observation and GIS databases, and web services. This is the front end application for the CUAHSI Hydrologic Information System Server. The HIS Server is a software stack that organizes observation databases, geographic data layers, data importing and management tools, and online user interfaces such as the DASH application, into a flexible multi- tier application for serving both national-level and locally-maintained observation data. The user interface of the DASH web application allows online users to query observation networks by location and attributes, selecting stations in a user-specified area where a particular variable was measured during a given time interval. Once one or more stations and variables are selected, the user can retrieve and download the observation data for further off-line analysis. The DASH application is highly configurable. The mapping interface can be configured to display map services from multiple sources in multiple formats, including ArcGIS Server, ArcIMS, and WMS. The observation network data is configured in an XML file where you specify the network's web service location and its corresponding map layer. Upon initial deployment, two national level observation networks (USGS NWIS daily values and USGS NWIS Instantaneous values) are already pre-configured. There is also an optional login page which can be used to restrict access as well as providing a alternative to immediate downloads. For large request, users would be notified via email with a link to their data when it is ready. http://river.sdsc.edu/dash/

H13H-1686 

Weather Downloader – A tool for downloading hydrologic data into Geographic Information Systems

* To, E S (eto@mail.utexas.edu), University of Texas at Austin, Center for Research in Water Resources University of Texas at Austin 10100 Burnet Road, Bldg 119, Austin, 78758, Whiteaker, T L (twhit@mail.utexas.edu), University of Texas at Austin, Center for Research in Water Resources University of Texas at Austin 10100 Burnet Road, Bldg 119, Austin, 78758, Maidment, D R (maidment@mail.utexas.edu), University of Texas at Austin, Center for Research in Water Resources University of Texas at Austin 10100 Burnet Road, Bldg 119, Austin, 78758,

The Consortium of Universities for the Advancement of Hydrologic Science (CUAHSI) is an organization aimed at developing infrastructure and services for the advancement of hydrologic science and education in the United States. A key development of CUAHSI is WaterOneFlow web services, which provide programmatic access to hydrologic data from a variety of data sources in a consistent manner, regardless of the data source. Weather Downloader is a Geographic Information System (GIS) application that uses WaterOneFlow web services to download the time series data needed to describe the hydrology of a given geographical area. It runs within ESRI™ 's ArcGIS environment and can access meteorological, groundwater and streamflow data from public repositories such as USGS and Daymet. In addition, it can be easily customized to access any source of data that have been published using the WaterOneFlow web service protocol. Weather Downloader interacts directly with ESRI™'s feature classes by systematically processing individual features and retrieving associated data into the TimeSeries table of an Arc Hydro data model (Maidment, 2002). Through the Arc Hydro data model, time series data can be integrated with spatial data and made accessible to GIS-based hydrologic and hydraulic simulation models.

Author(s) (2007), Title, Eos Trans. AGU, 88(52), Fall Meet. Suppl., Abstract #####-##.