HR: 14:10h
AN: H12G-03    [PDF]
TI: JUPITER Project - Joint Universal Parameter IdenTification and Evaluation of Reliability
AU: * Poeter, E
EM: epoeter@mines.edu
AF: Colorado School of MInes, 1516 Illinois St, Goden, CO 80401 United States
AU: Hill, M
EM: mchill@usgs.gov
AF: USGS, 3215 Marine St, Boulder, CO 80303 United States
AU: Doherty, J
EM: jdoherty@gil.com.au
AF: Watermark Computing and University of Queensland, 336 Cliveden Avenue, Brisbane, 4075 Australia
AU: Banta, E
EM: erbanta@usgs.gov
AF: USGS, Box 25046, Denver Federal Center, Lakewood, CO 80225-0046 United States
AU: Babendreier, J
EM: Babendreier.Justin@epamail.epa.gov
AF: USEPA, 960 College Station Road, Athens, GA 30605 United States
AB: The JUPITER (Joint Universal Parameter IdenTification and Evaluation of Reliability) project builds on the technology of two widely used codes for sensitivity analysis, data assessment, calibration, and uncertainty analysis of environmental models: PEST and UCODE. These programs are universal in that they can be applied to any computer model; and both have very flexible methods for interacting with application models through ASCII files. Their combination in an Application Programming Interface (API) will yield a full-featured, well-designed, flexible, stable, modular, thoroughly documented foundation for advancing the technology incorporated in UCODE and PEST. Phase 1 of the project is development of the JUPITER API, which will include (1) conventions for program input and output and internal data production and consumption, and (2) subroutines that support commonly used calculations and manipulations, to facilitate use of the API by many researchers in the field. Phase 2 is development of the first applications of the JUPITER API, J\_UCODE, J\_PEST, and J\_MMRI, where J\_MMRI is an alternative conceptual model evaluation tool for ranking and weighting models to facilitate multi-model inference. Applications developed using the JUPITER API will provide the opportunity for users to readily: (1) experiment with a number of techniques for generating conceptual models (e.g. geostatistical methods, geologic process modeling, upscaling); (2) compare alternative parameter-estimation algorithms (for example, the algorithms in J\_PEST and J\_UCODE); (3) "mine" results from various conceptual models for model evaluation, ranking and multi-model inferential analysis, as well as use these results to evolve the conceptual model (e.g. unreasonable parameter-value estimates provide clues to hydrogeologic structure; residual bias provides clues to conceptual model error); and (4) assess data needs to improve the calibration in light of the predictions.
DE: 1829 Groundwater hydrology
DE: 1832 Groundwater transport
SC: Hydrology [H]
MN: 2003 Fall Meeting