Show simple item record  

dc.contributor.authorJelley, Neilen_NZ
dc.date.accessioned2007-02-23T11:02:03Z
dc.date.available2007-08-21T16:38:07Z
dc.date.issued2007en_NZ
dc.identifier.citationJelley, N. (2007). Forecasting seasonal drawdowns in Whangamata town supply wells (Thesis, Master of Science (MSc)). The University of Waikato, Hamilton, New Zealand. Retrieved from https://hdl.handle.net/10289/2234en
dc.identifier.urihttps://hdl.handle.net/10289/2234
dc.description.abstractThe coastal township of Whangamata's reticulated water supply is provided by a number of groundwater bores, extracting water from local fractured rhyolite and andesite aquifers. A need has arisen to create a greater understanding of the aquifers, because of an increased demand for water abstraction. Water demand in Whangamata increases dramatically during the summer vacation period. Occupant numbers increase from 4,000 up to 50,000 during peak times, resulting in increased water demand. Over the past five years an increase in groundwater abstraction has produced an evident downward trend in bore water levels. Electrical conductivity is also increasing in several aquifers, posing a realistic threat of sea water intrusion and questioning the sustainability of current abstraction volumes. Multiple linear regression and an artificial neural network model were investigated as simple empirical forecasting tools for well drawdowns to predict the effect of future increases in groundwater demand. This approach was adopted as opposed to a groundwater numerical model because of poor time resolution of available data and the complex, fractured nature of the aquifer. By using pumping volumes as variables, seasonal bore water level variations and long term trends were predicted. The models were evaluated with independent validation data sets. The actual ability of a model to predict bore water level seasonal variation and long term trends was assessed using a comparison with a moving average of the validation data set. Multiple linear regression proved superior to the neural network in almost every bore modelled. Although neural networks proved capable of modelling seasonal bore water level variations it was not to the same degree of accuracy as the regression approach. The regression approach yielded a modified index of agreement of 0.6-0.74 when comparing a moving average of observed data with the validation data sets. The developed models were used to forecast well water levels with varying abstraction volumes aiming to prevent further long term decline in bore water levels.en_NZ
dc.format.mimetypeapplication/pdf
dc.format.mimetypeapplication/vnd.ms-excel
dc.format.mimetypeapplication/vnd.ms-excel
dc.language.isoen
dc.publisherThe University of Waikatoen_NZ
dc.rightsAll items in Research Commons are provided for private study and research purposes and are protected by copyright with all rights reserved unless otherwise indicated.
dc.subjectgroundwateren_NZ
dc.subjectcoastalen_NZ
dc.subjectsea water intrusionen_NZ
dc.subjectCoromandelen_NZ
dc.titleForecasting seasonal drawdowns in Whangamata town supply wellsen_NZ
dc.typeThesisen_NZ
thesis.degree.disciplineSchool of Science and Engineeringen_NZ
thesis.degree.grantorUniversity of Waikatoen_NZ
thesis.degree.levelMasters
thesis.degree.nameMaster of Science (MSc)en_NZ
uow.date.accession2007-02-23T11:02:03Zen_NZ
uow.date.available2007-08-21T16:38:07Zen_NZ
uow.identifier.adthttp://adt.waikato.ac.nz/public/adt-uow20070223.110203en_NZ
uow.date.migrated2009-06-09T23:31:30Zen_NZ
pubs.place-of-publicationHamilton, New Zealanden_NZ


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record