HOW TO analyse running weekly urtap solutions
or a series of running chopped urtap solutions
The idea behind this is to test the robustness of estimated
parameters as more and more data is coming in; for instance, does
the drift model converge?
WARNING! WORK IN PROGRESS
The working directory for running the analysis and the plot scripts
Running weekly tide analysis solutions are automatically saved in
The notion of result types, type1 type2
is presently inconsistent. We concieve of type1 as the
protocols (result sheets) above.
The significance is the representation (format) of the date of
analysis (and thus the end of the data window) in the file name.
type2 refers to experimental (re-processed)
Experimental results can be prepared with
which repeatedly calls cutts4urtap , extending the time
window of a urtap job further and further.
The resulting protocols and residual time series are saved in
If the instruction file (assigned by cutts4urtap) is changed, the
output directory should be moved
mv scut scut-your-choice
There are three scripts to plot the parameter evolution:
For plot-sol-evolution there
is a file with legend text that relates to the different four-letter
symbols signifying the time series in urtap's regression:
POLX Pol.Mot. cross-phase
POLI Pol.Mot. in-phase
BARO Local barometric admittance
ECCO Nontidal ocean
RIOS Kattegat sea level
Sa Annual tide
plot-drift-evolution -type2 -title 'S@-a@- tide
solved' -mark -cut -o ocut scut/*.prl
plot-tide-evolution -T 'Ssa Sta Mm Mf Mt' -title
'S@-a@- tide fixed at 1.16' -mark -cut-no-Sa -o ocut
-N 'POLI' -T Sa=ocut/Sa-evolamp-cut.dat -title 'S@-a@-
tide solved' -mark -cut -o ocut scut/
where the file in expression `Sa=ocut/Sa-evolamp-cut.dat´
has been produced by
plot-tide-evolution -T 'Sa Ssa Sta Mm Mf Mt'
-title 'S@-a@- tide solved' -mark -cut -o ocut scut/