1、Designation: D 5981 96 (Reapproved 2002)Standard Guide forCalibrating a Ground-Water Flow Model Application1This standard is issued under the fixed designation D 5981; the number immediately following the designation indicates the year oforiginal adoption or, in the case of revision, the year of las
2、t revision. A number in parentheses indicates the year of last reapproval. Asuperscript epsilon (e) indicates an editorial change since the last revision or reapproval.1. Scope1.1 This guide covers techniques that can be used tocalibrate a ground-water flow model. The calibration of amodel is the pr
3、ocess of matching historical data, and is usuallya prerequisite for making predictions with the model.1.2 Calibration is one of the stages of applying a ground-water modeling code to a site-specific problem (see GuideD 5447). Calibration is the process of refining the modelrepresentation of the hydr
4、ogeologic framework, hydraulicproperties, and boundary conditions to achieve a desireddegree of correspondence between the model simulations andobservations of the ground-water flow system.1.3 Flow models are usually calibrated using either themanual (trial-and-error) method or an automated (inverse
5、)method. This guide presents some techniques for calibrating aflow model using either method.1.4 This guide is written for calibrating saturated porousmedium (continuum) ground-water flow models. However,these techniques, suitably modified, could be applied to othertypes of related ground-water mode
6、ls, such as multi-phasemodels, non-continuum (karst or fracture flow) models, ormass transport models.1.5 Guide D 5447 presents the steps to be taken in applyinga ground-water modeling code to a site-specific problem.Calibration is one of those steps. Other standards have beenprepared on environment
7、al modeling, such as Guides D 5490,D 5609, D 5610, D 5611, D 5718, and Practice E 978.1.6 This standard does not purport to address all of thesafety concerns, if any, associated with its use. It is theresponsibility of the user of this standard to establish appro-priate safety and health practices a
8、nd determine the applica-bility of regulatory limitations prior to use.1.7 This guide offers an organized collection of informationor a series of options and does not recommend a specificcourse of action. This document cannot replace education orexperience and should be used in conjunction with prof
9、essionaljudgment. Not all aspects of this guide may be applicable in allcircumstances. This ASTM standard is not intended to repre-sent or replace the standard of care by which the adequacy ofa given professional service must be judged, nor should thisdocument be applied without consideration of a p
10、rojects manyunique aspects. The word “Standard” in the title of thisdocument means only that the document has been approvedthrough the ASTM consensus process.2. Referenced Documents2.1 ASTM Standards:D 653 Terminology Relating to Soil, Rock, and ContainedFluids2D 5447 Guide for Application of a Grou
11、nd-Water FlowModel to a Site-Specific Problem2D 5490 Guide for Comparing Ground-Water Flow ModelSimulations to Site-Specific Information2D 5609 Guide for Defining Boundary Conditions inGround-Water Flow Modeling2D 5610 Guide for Defining Initial Conditions in Ground-Water Flow Modeling2D 5611 Guide
12、for Conducting a Sensitivity Analysis for aGround-Water Flow Model Application2D 5718 Guide for Documenting a Ground-Water FlowModel Application2E 978 Practice for Evaluating Mathematical Models for theEnvironmental Fate of Chemicals33. Terminology3.1 Definitions:3.1.1 application verificationusing
13、the set of parametervalues and boundary conditions from a calibrated model toapproximate acceptably a second set of field data measuredunder similar hydrologic conditions.3.1.1.1 DiscussionApplication verification is to be distin-guished from code verification, which refers to softwaretesting, compa
14、rison with analytical solutions, and comparisonwith other similar codes to demonstrate that the code representsits mathematical foundations.3.1.2 calibrated modela model that has achieved a de-sired degree of correspondence between the model simulationsand observations of the physical hydrogeologic
15、system.3.1.3 calibration (model application)the process of refin-ing the model representation of the hydrogeologic framework,hydraulic properties, and boundary conditions to achieve a1This guide is under the jurisdiction of ASTM Committee D18 on Soil and Rockand is the direct responsibility of Subco
16、mmittee D18.21 on Ground Water andVadose Zone Investigations.Current edition approved July 10, 1996. Published November 1996.2Annual Book of ASTM Standards, Vol 04.08.3Annual Book of ASTM Standards, Vol 11.05.1Copyright ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19
17、428-2959, United States.desired degree of correspondence between the model simula-tions and observations of the ground-water flow system.3.1.4 calibration targetsmeasured, observed, calculated,or estimated hydraulic heads or ground-water flow rates that amodel must reproduce, at least approximately,
18、 to be consideredcalibrated.3.1.4.1 DiscussionThe calibration target includes both thevalue of the head or flow rate and its associated error ofmeasurement, so that undue effort is not expended attemptingto get a model application to closely reproduce a value whichis known only to within an order of
19、 magnitude.3.1.5 fidelitythe degree to which a model application isdesigned to resemble the physical hydrogeologic system.3.1.6 ground-water flow modelan application of a math-ematical model to represent a site-specific ground-water flowsystem.3.1.7 hydraulic propertiesproperties of soil and rock th
20、atgovern the transmission (for example, hydraulic conductivity,transmissivity, and leakance) and storage (for example, specificstorage, storativity, and specific yield) of water.3.1.8 inverse methodsolving for independent parametervalues using knowledge of values of dependent variables.3.1.9 residua
21、lthe difference between the computed andobserved values of a variable at a specific time and location.3.1.10 sensitivity (model application)the degree to whichthe model result is affected by changes in a selected modelinput representing hydrogeologic framework, hydraulic prop-erties, and boundary co
22、nditions.3.1.11 simulationin ground-water flow modeling, onecomplete execution of a ground-water modeling computerprogram, including input and output.3.2 For other definitions used in this guide, see TerminologyD 653.4. Summary of Guide4.1 The steps to be taken to calibrate a flow model are:establis
23、hing calibration targets and associated acceptable re-siduals or residual statistics (as described in Section 6),identifying calibration parameters (as described in Section 7),and history matching (as described in Section 8). Historymatching is accomplished by using the trial-and-error methodto achi
24、eve a rough correspondence between the simulation andthe physical hydrogeologic system, and then using either thetrial-and-error method or an automated method to achieve acloser correspondence.5. Significance and Use5.1 Most site-specific ground-water flow models must becalibrated prior to use in pr
25、edictions. In these cases, calibrationis a necessary, but not sufficient, condition which must beobtained to have confidence in the models predictions.5.2 Often, during calibration, it becomes apparent that thereare no realistic values of the hydraulic properties of the soil orrock which will allow
26、the model to reproduce the calibrationtargets. In these cases the conceptual model of the site mayneed to be revisited or the construction of the model may needto be revised. In addition, the source and quality of the dataused to establish the calibration targets may need to bereexamined. For exampl
27、e, the modeling process can sometimesidentify a previously undetected surveying error, which wouldresults in inaccurate hydraulic head targets.5.3 This guide is not meant to be an inflexible description oftechniques for calibrating a ground-water flow model; othertechniques may be applied as appropr
28、iate and, after dueconsideration, some of the techniques herein may be omitted,altered, or enhanced.6. Establishing Calibration Targets6.1 A calibration target consists of the best estimate of avalue of ground-water head or flow rate. Establishment ofcalibration targets and acceptable residuals or r
29、esidual statisticsdepends on the degree of fidelity proposed for a particularmodel application. This, in turn, depends strongly upon theobjectives of the modeling project. All else being equal, incomparing a low-fidelity to a high-fidelity model application,the low-fidelity application would require
30、 fewer calibrationtargets and allow larger acceptable residuals.NOTE 1Some low-fidelity models are not necessarily intended tomake specific predictions, but rather provide answers to speculative orhypothetical questions which are posed so as to make their predictionsconditional on assumptions. An ex
31、ample might be a model that answersthe question: “If the hydraulic conductivity of the soil is 50 feet per day,will the drawdown be more than 10 ft?” This model will not answer thequestion of whether or not the drawdown will, in reality, be more than 10ft because the value of hydraulic conductivity
32、was assumed. Since theanswer is conditional on the assumption, this “what-if” type of model doesnot necessarily require calibration, and, therefore, there would be nocalibration targets.6.2 For a medium- to high-fidelity model application, estab-lish calibration targets by first identifying all rele
33、vant availabledata regarding ground-water heads (including measured waterlevels, bottom elevations of dry wells, and top of casingelevations of flowing wells) and flow rates (including recordsof pumping well or wellfield discharges, estimates of baseflowto gaining streams or rivers or recharge from
34、losing streams,discharges from flowing wells, springflow measurements,and/or contaminant plume velocities). For each such datum,include the error bars associated with the measurement orestimate.6.3 Establish calibration targets before beginning any simu-lations.6.4 For any particular calibration tar
35、get, the magnitude ofthe acceptable residual depends partly upon the magnitude ofthe error of the measurement or estimate of the calibrationtarget and partly upon the degree of accuracy and precisionrequired of the models predictions. All else equal, the higherthe intended fidelity of the model, the
36、 smaller the acceptableabsolute values of the residuals.6.4.1 Head measurements are usually accurate to within afew tenths of a foot. Due to the many approximations em-ployed in modeling and errors associated therewith (see GuideD 5447), it is usually impossible to make a model reproduce allheads me
37、asurements within the errors of measurement. There-fore, the modeler must increase the range of acceptablecomputed heads beyond the range of the error in measurement.Judgment must be employed in setting these new acceptableresiduals. In general, however, the acceptable residual shouldbe a small frac
38、tion of the difference between the highest andD 5981 96 (2002)2lowest heads across the site.NOTE 2Acceptable residuals may differ for different hydraulic headcalibration targets within a particular model. This may be due to differenterrors in measurement, for example, when heads at some wells are ba
39、sedon a survey, but other heads are estimated based on elevations estimatedfrom a topographic map. In other circumstances, there may be physicalreasons why heads are more variable in some places than in others. Forexample, in comparing a well near a specified head boundary with a wellnear a ground-w
40、ater divide, the modeled head in the former will dependless strongly upon the input hydraulic properties than the head in the latter.Therefore, acceptable residuals near specified head boundaries can be setlower than those near divides.NOTE 3One way to establish acceptable hydraulic head residuals i
41、s touse kriging on the hydraulic head distribution. Although kriging is notusually recommended for construction of hydraulic head contours, it doesresult in unbiased estimates of the variance (and thus standard deviation)of the hydraulic head distribution as a function of location within themodeled
42、domain. The acceptable residual at each node can be set as thestandard deviation in the hydraulic head at that location. Some researchersquestion the validity of this technique (1).4An alternative is to performtrend analysis of regions of similar heterogeneity. Since a model willusually only be able
43、 to represent trends over length scales larger than thescale of local heterogeneity that is causing variations, the magnitude of theresiduals from the trend analysis should approximate the magnitude ofresiduals in the model in that region.6.4.2 Errors in the estimates of ground-water flow rates will
44、usually be larger than those in heads (2). For example,baseflow estimates are generally accurate only to within anorder of magnitude. In such cases, the upper and lower boundson the acceptable modeled value of baseflow can be equal tothe upper and lower bounds on the estimate.6.5 Multiple Hydrologic
45、 ConditionsWhen more than oneset of field measurements have been collected, identify thedifferent hydrologic conditions that are represented by theavailable data sets. Include only one data set from eachhydrologic condition in the set of calibration targets. Use theremaining data sets for verificati
46、on.6.5.1 Uniqueness (Distinct Hydrologic Conditions)Thenumber of different distinct hydrologic conditions that a givenset of input aquifer hydraulic properties is capable of repre-senting is an important qualitative measure of the performanceof a model. It is usually better to calibrate to multipleh
47、ydrologic conditions, if the conditions are truly distinct.Matching different hydrologic conditions is one way to addressnonuniqueness, because one set of heads can be matched withthe proper ratio of ground-water flow rates to hydraulicconductivities; whereas, when the flow rates are changed,represe
48、nting a different condition, then the range of hydraulicconductivities that produce acceptable residuals becomes muchmore limited.6.5.1.1 Other ways to address the uniqueness problem are toinclude ground-water flows with heads as calibration targets,and to use measured values of hydraulic properties
49、 as modelinputs.6.5.2 Verification (Similar Hydrologic Conditions)Whendata are available for two times of similar hydrologic condi-tions, only one of those data sets should be used as calibrationtargets because they are not distinct. However, the other dataset can be used for application verification. In the verificationprocess, the modeled data are compared, not to the calibrationdata set, but to the verification data set. The resulting degree ofcorrespondence can be taken as an indicator or heuristicmeasure of the uncertainty inherent in the models predictions.NO
