Design of efficient groundwater monitoring networks
Thesis/Dissertation
·
OSTI ID:6565122
This study attacks the problem of designing minimum groundwater monitoring networks satisfying accuracy requirements. The design should include five common variables: parameters to sample, number of stations, location of stations, sampling frequency and duration of the sampling program. This study is concerned with the efficient determination of the second and third variables, number and location of sampling stations using the theory of regionalized variables. Traditionally, selecting the number and location of sampling stations have relied on the designer's experience or on the decision maker's subjective evaluation of the program, both working under some budgetary constraint. This study approaches the problem using an accuracy of measurement criterion (kriging variance) to best select the variables. The theory of regionalized variables uses the universal kriging algorithm, that is an unbiassed minimum variance estimator capable of providing the mean square error of estimation, as a measurement of accuracy. Two cases were considered; the first reinforcing an existing groundwater monitoring network, where Kriging shows if it is advisable to reinforce the system. Then, a gradient search is constructed to select the optimal location of the additional station. The second case considers the design of new monitoring networks, where a procedure was developed to obtain the network which provides the minimum variance of the estimate for an increasing number of monitoring stations. The parameters for the kriging estimation were obtained from groundwater quality data that originated in a variety of geohydrological conditions. In all the cases studied the variance of estimation converge to minimum values rapidly, suggesting that only a few monitoring wells were sufficient to yield good estimates.
- OSTI ID:
- 6565122
- Country of Publication:
- United States
- Language:
- English
Similar Records
Evaluating techniques for multivariate classification of non-collocated spatial data.
Regionalized variable theory and assessment of the consequences of uncertainty in optimal design of water-quality sampling networks
A kriging and entropy-based approach to raingauge network design
Conference
·
Wed Sep 01 00:00:00 EDT 2004
·
OSTI ID:972884
Regionalized variable theory and assessment of the consequences of uncertainty in optimal design of water-quality sampling networks
Thesis/Dissertation
·
Tue Dec 31 23:00:00 EST 1985
·
OSTI ID:5282291
A kriging and entropy-based approach to raingauge network design
Journal Article
·
Wed Feb 14 23:00:00 EST 2018
· Environmental Research
·
OSTI ID:23100690