Modeling - Field Sampling: Principles and Practices in Environmental Analysis - page 184

MINTEQ Femy AR, Girivin DC, Jenne EA. MINTEQ: A Computer Program for Calculating

Aqueous Geochemical Equilibria. NTIS PB84-157148, EPA-600/3-84-032.
Springfield, VA: National Technical Information Service, 1984.

C-SALT Smith GR, Tanji KK, Burau JJ. C-SALT—A Chemical Equilibrium Model for

Multicomponent Solutions. Loeppert RH, et al., eds. SSSA spec. pub. 42. Madison,
WI: ASA and SSSA, 1990.

FITEQL Westall JC, Morel EMM. FITEQL: A General Algorithm for the Determination of

Metal Ligand Complex Stability Constants from Experimental Data. tech. note 19.
Cambridge, MA: Ralph M. Parson Lab., Dept. Civil Eng., 1977.

Note: ASA, Agronomy Society of America; SSSA, Soil Science Society of America.

of a model may not be apparent until it is in use. The second step is calibration of the
model.
The third step is validation, which must be done in all cases; that is, if the model
actually predicts what will happen in the environment. Verification is usually done by
comparison of a model to a field situation. In restricted areas with uniform or very similar
soils models can often accurately predict the form and movement of a contaminant in the
environment. Caution is important, however. Often the model will not be as effective or
may not predict the field situation at all when it is applied to field situations in other
localities; thus for each new application of a model a validation experiment must be done.
This is especially true if the model is to be used to predict sampling site locations.
Most computer models contain empirical components that are varied by the researcher
to produce an acceptable result. Some models contain several such factors related to
various components. Another approach is to use weighting factors. These can be used to
increase or decrease the importance of a particular component of the equation so that the
results better correspond to what is actually happening in the environment.
All of these types of models suffer from the fact that they need large, sometimes vast,
arrays of inputs and variables to make them useful. In many cases these variables and
constants are not known. In these cases they can be obtained either by making field
measurements (measuring a soil's porosity, bulk density, etc.) or using column and
lysimeter studies. Another approach is to estimate by various means the numbers from
other known characteristics of the soil or landscape. This is often done during a
calibration stage. For example, in groundwater (GW) modeling, well pumping, along
with pressure head measurements, might be taken to calibrate a GW model. In this case,
calibration is done by, for example, adjusting the value for soil conductivity. Once
calibration is completed other data are used for the verification phase. After the
verification phase and once all the data are collected or estimated a calculation can be
made.

7.3. SIMPLE MATHEMATICAL MODELS

A number of soil samples from differing soil types, specifically different soil orders from