skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: FUNCTIONAL SENSITIVITY ANALYSIS FOR COMPUTER MODEL OUTPUT

Abstract

No abstract prepared.

Authors:
Publication Date:
Research Org.:
Los Alamos National Lab., NM (US)
Sponsoring Org.:
US Department of Energy (US)
OSTI Identifier:
790205
Report Number(s):
LA-UR-02-14
TRN: US200306%%419
DOE Contract Number:
W-7405-ENG-36
Resource Type:
Conference
Resource Relation:
Conference: Conference title not supplied, Conference location not supplied, Conference dates not supplied; Other Information: PBD: 1 Jan 2002
Country of Publication:
United States
Language:
English
Subject:
99 GENERAL AND MISCELLANEOUS//MATHEMATICS, COMPUTING, AND INFORMATION SCIENCE; COMPUTERS; FUNCTIONALS; SENSITIVITY ANALYSIS

Citation Formats

K. Campbell. FUNCTIONAL SENSITIVITY ANALYSIS FOR COMPUTER MODEL OUTPUT. United States: N. p., 2002. Web.
K. Campbell. FUNCTIONAL SENSITIVITY ANALYSIS FOR COMPUTER MODEL OUTPUT. United States.
K. Campbell. Tue . "FUNCTIONAL SENSITIVITY ANALYSIS FOR COMPUTER MODEL OUTPUT". United States. doi:. https://www.osti.gov/servlets/purl/790205.
@article{osti_790205,
title = {FUNCTIONAL SENSITIVITY ANALYSIS FOR COMPUTER MODEL OUTPUT},
author = {K. Campbell},
abstractNote = {No abstract prepared.},
doi = {},
journal = {},
number = ,
volume = ,
place = {United States},
year = {Tue Jan 01 00:00:00 EST 2002},
month = {Tue Jan 01 00:00:00 EST 2002}
}

Conference:
Other availability
Please see Document Availability for additional information on obtaining the full-text document. Library patrons may search WorldCat to identify libraries that hold this conference proceeding.

Save / Share:
  • No abstract prepared.
  • Computer-models which attempt to define interactions among dynamic parameters believed to influence the development of ''cyclic'' carbonate platform sequences have been popularized over the past few years. These models typically utilize vectors for subsidence (constant) and cyclical (sinusoidal) eustatic sea-level to create accommodation space which is filled by sedimentation (depth-dependent rates) following an appropriate lag time (non-depositional episode during initial platform flooding). Since these models are intended to reflect general principles of cyclic carbonate deposition, it is instructive to test their predictive utility by comparing typical model outputs with an actively evolving depositional cycle on a modern carbonate platform wheremore » rates of subsidence, eustatic sea-level and sediment accumulation are known. Holocene carbonate deposits across northern Great Bahama Bank provide such an ideal test-platform for model-data comparisons. On Great Bahama Bank, formation of accommodation space depends on eustatic sea-level rise because tectonic subsidence is very slow. Contrary to typical model input parameters, however, the rate of formation of accommodation space varies irregularly across the bank-top because irregular bank-top topography (produced by subaerial erosion and karstification) results in differential flooding of the platform surface. Results of this comparison indicate that typical computer-model input variables (subsidence, sea-level, sedimentation, lag-time) and output depositional geometries are poorly correlated with real depositional patterns across Great Bahama Bank. Since other modern carbonate platforms and ancient carbonate sequences display similarly complex stratigraphies, it is suggested that present computer-modeling results have little predictive value for stratigraphic interpretation.« less
  • An automated procedure for performing sensitivity analyses has been developed. The procedure uses a new FORTRAN compiler with computer calculus capabilities to generate the derivatives needed to set up sensitivity equations. The new compiler is called GRESS - Gradient Enhanced Software System. Application of the automated procedure with ''direct'' and ''adjoint'' sensitivity theory for the analysis of non-linear, iterative systems of equations is discussed. Calculational efficiency consideration and techniques for adjoint sensitivity analysis are emphasized. The new approach is found to preserve the traditional advantages of adjoint theory while removing the tedious human effort previously needed to apply this theoreticalmore » methodology. Conclusions are drawn about the applicability of the automated procedure in numerical analysis and large-scale modelling sensitivity studies. 24 refs., 2 figs.« less