The WS Trend Removal Module is a useful supplement to the WS Digital Filtering Analysis Module. Hence, the traditional filtering cannot solve some simple problems, as the duration of a trend (which is infinite) is much longer than the duration of the filter (or time series).
One example is a measured surface elevation with a very slow, unilateral variation superimposed (e.g. due to a drifting sensor or leakage from a wave flume). The removal of the superimposed variation is accomplished by approximating the time series by an orthogonal polynomial fit of lower order that subsequently is subtracted from the time series.
Figure 9.1 The upper figure shows a measured time series with a unilateral variation due to a drifting gauge. The lower figure shows the detrended time series and the removed trend.
