Daily values are obtained using a two-step filtering
operation.  First, the dominant diurnal and semi-diurnal tidal
components are removed from the quality controlled hourly values.
Secondly, a 119-point convolution filter (Bloomfield, 1976)
centered on noon is applied to remove the remaining high
frequency energy and to prevent aliasing when the data are
computed to daily values.  The 95, 50, and 5% amplitude points
are 124.0, 60.2, and 40.2 hours, respectively.  The Nyquist
frequency of the daily data is at a period of 48 hours which has
a response of about 5% amplitude, thus, aliasing is minimal.  The
primary tidal periods have a response of less than 0.1%
amplitude.

     The filtering operation incorporates an objective procedure
to handle gaps.  This objective technique simply replaces the
filter weight at any missing observation  with a zero and
renormalizes the sum of the modified weight function to unity.
This technique is equivalent to interpolating the missing
observation with an estimate of the local mean of the time
series.  The local mean is defined as the mean of a given segment
of length equal to the length of the filter.  The error
associated with this technique can be estimated objectively and
is used as a criterion for accepting or rejecting a daily value
computed in an area of the time series which contains a gap or
gaps.  This error depends on the ratio of the standard deviations
of the input (hourly) and the output (daily) data.  Thus in order
to keep the ratio low, it is essential to apply this technique to
the residual series as defined above.

     The monthly values are calculated from the daily data with a
simple average of all the daily values in a month. If seven or
fewer values are missing, the monthly value is calculated.