HR: 0800h
AN: H31G-0738    [Abstracts]
TI: Evaluation of Alternative Conceptual Models Using Interdisciplinary Information: An Application in Shallow Groundwater Recharge and Discharge
AU: * Lin, Y
EM: yflin@uiuc.edu
AF: Illinois State Water Survey and University of Illinois at Urbana-Champaign, 2204 Griffith Dr., Champaign, IL 61820,
AU: Bajcsy, P
EM: pbajcsy@ncsa.uiuc.edu
AF: National Center for Supercomputing Applications and University of Illinois at Urbana- Champaign, 1205 W. Clark St., Urbana, IL 61801,
AU: Valocchi, A J
EM: valocchi@uiuc.edu
AF: Department of Civil and Environmental Engineering, University of Illinois at Urbana- Champaign, 205 North Mathews Ave., Urbana, IL 61801,
AU: Kim, C
EM: ckim5@uiuc.edu
AF: National Center for Supercomputing Applications and University of Illinois at Urbana- Champaign, 1205 W. Clark St., Urbana, IL 61801,
AU: Wang, J
EM: jwang41@uiuc.edu
AF: Department of Civil and Environmental Engineering, University of Illinois at Urbana- Champaign, 205 North Mathews Ave., Urbana, IL 61801,
AB: Natural systems are complex, thus extensive data are needed for their characterization. However, data acquisition is expensive; consequently we develop models using sparse, uncertain information. When all uncertainties in the system are considered, the number of alternative conceptual models is large. Traditionally, the development of a conceptual model has relied on subjective professional judgment. Good judgment is based on experience in coordinating and understanding auxiliary information which is correlated to the model but difficult to be quantified into the mathematical model. For example, groundwater recharge and discharge (R&D) processes are known to relate to multiple information sources such as soil type, river and lake location, irrigation patterns and land use. Although hydrologists have been trying to understand and model the interaction between each of these information sources and R&D processes, it is extremely difficult to quantify their correlations using a universal approach due to the complexity of the processes, the spatiotemporal distribution and uncertainty. There is currently no single method capable of estimating R&D rates and patterns for all practical applications. Chamberlin (1890) recommended use of "multiple working hypotheses" (alternative conceptual models) for rapid advancement in understanding of applied and theoretical problems. Therefore, cross analyzing R&D rates and patterns from various estimation methods and related field information will likely be superior to using only a single estimation method. We have developed the Pattern Recognition Utility (PRU), to help GIS users recognize spatial patterns from noisy 2D image. This GIS plug-in utility has been applied to help hydrogeologists establish alternative R&D conceptual models in a more efficient way than conventional methods. The PRU uses numerical methods and image processing algorithms to estimate and visualize shallow R&D patterns and rates. It can provide a fast initial estimate prior to planning labor intensive and time consuming field R&D measurements. Furthermore, the Spatial Pattern 2 Learn (SP2L) was developed to cross analyze results from the PRU with ancillary field information, such as land coverage, soil type, topographic maps and previous estimates. The learning process of SP2L cross examines each initially recognized R&D pattern with the ancillary spatial dataset, and then calculates a quantifiable reliability index for each R&D map using a supervised machine learning technique called decision tree. This JAVA based software package is capable of generating alternative R&D maps if the user decides to apply certain conditions recognized by the learning process. The reliability indices from SP2L will improve the traditionally subjective approach to initiating conceptual models by providing objectively quantifiable conceptual bases for further probabilistic and uncertainty analyses. Both the PRU and SP2L have been designed to be user-friendly and universal utilities for pattern recognition and learning to improve model predictions from sparse measurements by computer-assisted integration of spatially dense geospatial image data and machine learning of model dependencies.
UR: http://isda.ncsa.uiuc.edu/download/
DE: 1819 Geographic Information Systems (GIS)
DE: 1830 Groundwater/surface water interaction
DE: 1847 Modeling
DE: 1880 Water management (6334)
DE: 1894 Instruments and techniques: modeling
SC: Hydrology [H]
MN: 2007 Fall Meeting