--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/Documentation/README.procfigs	Fri Mar 09 08:34:30 2012 +0000
@@ -0,0 +1,80 @@
+======================================================================
+                    R E A D M E . P R O C F I G S 
+                    doc: Wed Feb 15 18:43:24 2012
+                    dlm: Wed Feb 15 19:15:40 2012
+                    (c) 2012 A.M. Thurnherr
+                    uE-Info: 71 0 NIL 0 0 72 3 2 4 NIL ofnI
+======================================================================
+
+During processing of, say station 067, several figures are produced.
+Where applicable, downcast data are colored orange and upcast data in
+green. Red/blue of signed data indicate positive/negative values.
+
+=067_prof.eps=
+
+This figure shows the the final result of processing and contains three
+plots: 
+
+1) binned vertical velocity profiles (thick solid lines, lower x axis)
+2) mean-absolute-deviation in each bin (bullets, lower x axis)
+3) number of samples in each bin (thin solid lines, upper x axis)
+
+
+=067_w.eps=
+
+This figure shows a depth-time-series of the vertical velocities. Under
+most circumstances, real vertical-velocity features will have a
+horizontally banded structure --- for example, an interface between
+downwelling and upwelling (a blue-red transition) is not expected to
+move much vertically during the time it is observed by the ADCP. For
+casts without significant vertical velocities, this plot should show
+random noise.
+
+
+=067_residuals.eps=
+
+This figure shows a depth-time-series of vertical-velocity residuals,
+defined as
+
+residual(ensemble,bin) = w_measured(ensemble,bin) - w_profile(depth(ensemble,bin))
+
+This plot should show random noise.
+
+
+=067_BR.eps=
+
+In this figure, mean residuals are plotted as a function of bin number
+(i.e. distance from the ADCP). The profiles should be near zero,
+especially at low bin numbers.
+
+
+=067_Sv.eps=
+
+This figure shows a depth-time-series of acoustic volume scattering
+coefficient, following Deines (1999). The very apparent trend with
+distance from the ADCP implies that this method does not work very
+well. Nevertheless, this plot can be useful, in particular to detect
+erroneous vertical velocities associated with diel migration of
+zooplankton.
+
+
+=067_corr.eps=
+
+This figure shows a depth-time-series of correlation. It is probably not
+useful.
+
+
+=067_TL.eps=
+
+This figure shows the optimal time lags, calculated in small-ish
+windows, as a function of time. Note that each scan is 1/24th of a
+second. Outliers do not matter, but jumps and significant trends
+indicate dropped CTD scans (due to transmission problems between the
+CTD and the deck box). Such casts should be treated with skepticism.
+
+
+=067_TLhist.eps=
+
+This figure shows a histogram of the optimal time lags. It is probably
+not useful.
+