PERMFILE42740

• Permit No: M2001051
• IBM Index Class Name: Permit File
• Doc Date: 5/31/2001
• Doc Name: HOME OFFICE WELL MONITORING PROTOCOL PROJECT 864-050
• From: TST
• To: WESTERN MOBILE NORTHERN LAFARGE
• Media Type: D
• Archive: No

Appendix C for this purpose, including the computer file. Data must be transferred from
the data log sheets (or spreadsheet) to the specially formatted spreadsheet file that will be
used by the software. This should ideally be done at one time, by one person, at the time
analysis will be performed, to ensure consistent data entry techniques. A final check of
data should confirm, at minimum, that for each well, no measurements exceed the
' originally measured depth of the well. Such measurements, or others with similarly
suspicious values should be removed from the data set. At this time the single highest
and lowest measured values will simply be removed from the input file and replaced with
' a "n/a" which signals the program that no measurement was taken. This holds the date's
place but removes the data from the analysis. These removals should not be done on
either the original data log or the data log computer files, but only in the WQSTAT input
' file.
Since this analysis will not be completed for at least 3-4 years from the start of
' monitoring activities, it is recommended Lafarge retain all the necessary software and
manuals for analyzing the data collected. Although many packages are available, this
program has been written specifically for water quality statistical analysis and is
' exceptionally user friendly. Even with the best of intentions, personnel changes and
other unforeseen events can happen in that period of time. Precautions should be taken to
insure the needed information can be produced once monitoring has been completed.
'
A few notes on procedure for WQSTAT Plus may be in order. The input file will include
a column labeled "Flow" that will be used by the software to translate the measurements
' to flow-adjusted values. The flow adjustment calculation is run from the "Examine Data"
feature. WQSTAT adjusts data for flow using slog-log relationship as follows:
Log (groundwater level) = b(log flow) + a
WQSTAT uses linear regression to estimate the slope and intercept of the line above.
Then from each observation (log groundwater level) the corresponding prediction based
on flow, b(log flow) + a, is subtracted, producing a series of residuals with a mean of
' zero. To each residual, the mean of the original groundwater level series is added,
producing a flow adjusted series of log groundwater levels which has the same mean as
the original. Then the antilogs of the log groundwater levels are found and a final
correction is made so the resulting series will have the same mean as the original series of
observations.
Once the adjustment is made, to use the flow adjusted data the "Use Alternate Data" box
must be checked in the "Examine Data" feature. If this box is left unchecked the original
measured data will be used. This may be a useful comparison to indicate how strongly
groundwater levels are correlated to river flow.
WQSTAT allows use of parametric, non-parametric, one- and two-tailed tolerance limit
' methods. As mentioned above, the non-parametric two-tailed method should be selected
for this study
.

' 11