A Neural‐Network Approach to the Determination of Aquifer Parameters

Abd Rashid Abd Aziz, Kau‐Fui Vincent Wong

Research output: Contribution to journalArticle

78 Scopus citations

Abstract

A new approach to determine aquifer parameter values from aquifer-test data has been developed that uses the pattern-matching capability of a neural network. The network is trained to recognize patterns of normalized drawdown data as input and the corresponding aquifer parameters as output. The Theis and Hantush-Jacob solutions for confined and leaky-confined aquifer conditions are used to derive the input patterns based on the parameter values selected from predetermined ranges. The trained network produces output of aquifer parameter values when it receives the aquifer-test data as the input patterns. The results obtained from this new approach are in good agreement with published results using other techniques. The advantages of the present approach are the automated process of obtaining aquifer parameter values and the ability of the network to associate drawdown to the corresponding Theis and Hantush-Jacob solutions.

Original languageEnglish (US)
Pages (from-to)164-166
Number of pages3
JournalGroundwater
Volume30
Issue number2
DOIs
StatePublished - Mar 1992

ASJC Scopus subject areas

  • Water Science and Technology
  • Computers in Earth Sciences

Fingerprint Dive into the research topics of 'A Neural‐Network Approach to the Determination of Aquifer Parameters'. Together they form a unique fingerprint.

  • Cite this