>>>> INSTRUCTION: num-instruction-files (NO DELETE)
iTOUGH2 capabilities can be applied to non-TOUGH2 models using the PEST interface. Instruction files are used to communicate between the output variables of the user-supplied model (which must be provided through one or multiple ASCII text files) and the iTOUGH2 observation vector. (Note that if the model writes output to the screen, the “>” symbol can be used on the command line (see >>>> EXECUTABLE) to redirect standard output from the screen to a text file.)
An instruction file must be provided for each output file that contains an observable variable used by iTOUGH2 for model evaluation; num-instruction-files is the number of instruction files provided. Instruction files are matched to their corresponding output files on the lines following the >>>> INSTRUCTION command.
Upon initialization, all output files are deleted unless keyword NO DELETE is present.
PEST requires that at least one PEST observation is provided. However, if the external model is a preprocessor to TOUGH2, all observations may be internally taken from TOUGH2 arrays, without the need for an observation that is read from an external output file. Dummy instruction and output files can be automatically generated by providing their respective file names as dummy.ins and dummy.out. A “measured” value of zero will be generated and returned for comparison to an observation named dummy that needs to be defined in the iTOUGH2 input file.
Under Unix, it is recommended to add separate lines with the keyword FILE: followed by the file name, so the files are automatically copied to the temporary directory. This can occur anywhere in the iTOUGH2 input file.
>>>> EXECUTABLE : "idratherusetough.exe > wadyaexpect"
>>>> TEMPLATE : 1
>>>> INSTRUCTION : 2
copy FILE: input.tpl to temporary directory
copy FILE: output1.ins to temporary directory
copy FILE: output2.ins to temporary directory
copy FILE: idratherusetough.exe to temporary directory