496 CHAPTER 8. LABORATORY MEASUREMENT AND ANALYSIS
and trends are easily removed from the data if called for
by the intended analysis. Trend removal should only be
performed if the trend is physically expected or clearly
apparent in the data (Bendat and Piersol 1971).
3. Time series wave records are sometimes “filtered” to remove unwanted higher or lower frequency components before continuing with subsequent analyses. For example,
if the primary interest is in the second-order bound long
wave, the shorter waves can be removed from the signal by
a “low-pass filter.” Similarly, lower frequency fluctuations
can be removed using a “high-pass filter.”
Whether the data editing is performed automatically or under human
supervision, the goal is to minimize any negative effects the editing might
have on the veracity of the measured data. In addition, it is wise to save the
original unedited records for possible future use that may require a different
editing procedure.
8.3.2 Time Domain (Statistical) Analysis
Time domain analysis provides representative statistical descriptors of the
wave data that are useful in coastal design, particularly in developing empirical relationships between important parameters used in design guidance.
Time domain analysis includes nonlinear wave effects in a more direct manner than the linear techniques used in spectral analysis (Thompson and
Long 1987) . Consequently, this method of analysis is important for shallow water waves. Time domain analysis is used primarily on records of sea
surface elevation variations; however, the methods can also be applied to
any time-varying data records such as wave orbital velocities or pressures.
Usually, the first step in analyzing time series wave data is to calculate
the mean water level and subtract it from the time series. Also, this is the
time to remove any known linear trends in the data that were not been
removed during data editing. The resulting “zero-meaned” time series is
ready for analysis.
Two types of time domain wave statistics can be obtained from times
series data. The first type uses “TV” evenly spaced values of sea surface
elevation (rj,) to calculate statistics related to the sea surface elevations.
Typically calculated are the following statistics (Goda 1985):
and trends are easily removed from the data if called for
by the intended analysis. Trend removal should only be
performed if the trend is physically expected or clearly
apparent in the data (Bendat and Piersol 1971).
3. Time series wave records are sometimes “filtered” to remove unwanted higher or lower frequency components before continuing with subsequent analyses. For example,
if the primary interest is in the second-order bound long
wave, the shorter waves can be removed from the signal by
a “low-pass filter.” Similarly, lower frequency fluctuations
can be removed using a “high-pass filter.”
Whether the data editing is performed automatically or under human
supervision, the goal is to minimize any negative effects the editing might
have on the veracity of the measured data. In addition, it is wise to save the
original unedited records for possible future use that may require a different
editing procedure.
8.3.2 Time Domain (Statistical) Analysis
Time domain analysis provides representative statistical descriptors of the
wave data that are useful in coastal design, particularly in developing empirical relationships between important parameters used in design guidance.
Time domain analysis includes nonlinear wave effects in a more direct manner than the linear techniques used in spectral analysis (Thompson and
Long 1987) . Consequently, this method of analysis is important for shallow water waves. Time domain analysis is used primarily on records of sea
surface elevation variations; however, the methods can also be applied to
any time-varying data records such as wave orbital velocities or pressures.
Usually, the first step in analyzing time series wave data is to calculate
the mean water level and subtract it from the time series. Also, this is the
time to remove any known linear trends in the data that were not been
removed during data editing. The resulting “zero-meaned” time series is
ready for analysis.
Two types of time domain wave statistics can be obtained from times
series data. The first type uses “TV” evenly spaced values of sea surface
elevation (rj,) to calculate statistics related to the sea surface elevations.
Typically calculated are the following statistics (Goda 1985):