Calibration of models using groundwater age

There have been substantial efforts recently by geo-chemists to determine the age of groundwater (time sincewater entered the system) and its uncertainty, and byhydrologists to use these data to help calibrate ground-water models. This essay discusses the calibration ofmodels using groundwater age, with conclusions thatemphasize what is practical given current limitationsrather than theoretical possibilities.From a research perspective, it can be argued that theconservation equations imposed in groundwater modelsmake them most useful for providing fundamental insightsinto how systems operate. Yet from a water-managerperspective, water-use decisions must be made, and thusmodels are created as mathematical representations ofaquifer systems in order to compile all of the availableinformation into a tool that honors the physics andchemistry of the system. The compiled information willinclude prior knowledge of hydrogeologic frameworkparameters, but as the values and spatial distribution ofthose parameters are never completely known, hydraulicdata are used to further constrain them during calibration.These hydraulic data can be divided into two categories:potentials (heads or water levels) and fluxes (directmeasurements or indirect estimates). Having only headvalues will result in a nonunique answer with no optimalsolution; flux information of some kind (whether pre-scribed or estimated) must be included during calibration.Direct measurements of fluxes could include those ofrecharge, baseflow, or well pumpage. In many geographicregions (arid climates, for example), data or estimates forthe first two are often not available. Pumpage data arefrequently available, but the flow is often small comparedto regional flow and, thus, although adding pumpage datamay reduce nonuniqueness, large uncertainties in theestimated parameters often remain. Indirect estimates offluxes could include those of recharge (using water budgetmethods) or environmental tracers. Tracers include datasuch as temperature, chemical concentrations, or apparentage interpreted from the chemistry. Age is proving to be awidely used (Kazemi et al. 2006; Bethke and Johnson2008) flux indicator, although it is not independent of theeffective porosity, which will be discussed shortly.