Explaining long-term trends in groundwater hydrographs

R. Ferdowsian, David Pannell

    Research output: Chapter in Book/Conference paperConference paperpeer-review

    15 Citations (Scopus)

    Abstract

    An ability to understand and interpret changes in groundwater levels is essential for sound management of groundwater resources. Various statistical methods have been developed to explain hydrograph trends. Most of these operate on the assumption that groundwater trends are linear or best represented by short linear segments. However there is clear evidence that many hydrograph trends are nonlinear. For example, as the groundwater level changes, the area of groundwater discharge may change, dynamically feeding back and altering the trend rate of change. The long-term underlying trend of groundwater level over time may or may not be linear. Three types of long-term trends may be observed: • We may observe a rising trend that reduces over time and is a common feature of local groundwater flow systems. In these cases, as groundwater level rise, the hydraulic gradient and the rate of flow (discharge) both increase. The result is that the rate of groundwater rise falls systematically over time. • There are cases where the long-term trend in groundwater level is linear. This is usually the case where aquifers are intermediate to regional, have relatively higher recharge to discharge ratios, very little hydraulic gradient to generate significant flow and groundwater levels are well-below the soil surface. • Occasionally we may observe an increasing rate of groundwater rise. This may be observed where stagnant aquifers are segmented by obstacles (eg basement highs). In such cases each segment will rise until the area of discharge has increased or groundwater finds a convenient flow path to spill into the lower part of the aquifer. From that time, the lower segment, receiving water from another segment, will have an increasing rate of groundwater rise, at least for a period. We present three models for statistically estimating non-linear trends in groundwater levels. The first model is an autoregressive model, in which a past moving average of the dependant variable is included as an explanatory variable. This approach is useful when regular and frequent water-level data is available, although it has a few shortcomings. The second model uses time-related spline functions in the GenStat statistics package. The third model includes a log time function to capture the non-linear trend. Both, the spline and log time models are easy to use and produce realistic trend lines. The spline function is a subjective method as the number of splines needs to be selected. The advantage of the log time model over the spline method is that it is not subjective and does not need special software; the solver function of Microsoft Excel may be used to do the analysis. We provide detailed guidance on performing the spline and log time models. © MODSIM 2009.All rights reserved.
    Original languageEnglish
    Title of host publicationThe 18th World IMACS Congress and MODSIM09 International Congress on Modelling and Simulation
    Editors Anderssen, R.S., R.D. Braddock, L.T.H. Newham
    Place of PublicationAustralia
    PublisherModelling and Simulation Society of Australia and New Zealand Inc.
    Pages3060-3066
    VolumeMODSIM09
    EditionCairns, Australia
    ISBN (Print)9780975840078
    Publication statusPublished - 2009
    Event18th World IMACS Congress and International Congress on Modelling and Simulation: Interfacing Modelling and Simulation with Mathematical and Computational Sciences, MODSIM 2009 - Cairns, Australia
    Duration: 13 Jul 200917 Jul 2009

    Conference

    Conference18th World IMACS Congress and International Congress on Modelling and Simulation: Interfacing Modelling and Simulation with Mathematical and Computational Sciences, MODSIM 2009
    Country/TerritoryAustralia
    CityCairns
    Period13/07/0917/07/09

    Fingerprint

    Dive into the research topics of 'Explaining long-term trends in groundwater hydrographs'. Together they form a unique fingerprint.

    Cite this