Definitions of Time-Series Outliers

I am using SPSS's Expert Modeler Option to Model a ARIMA model.

I see that there is an option to have SPSS automatically detect various type of outliers.

I get what the following outliers look like:
  • Additive
  • Level Shift
  • Transient
  • Seasonal Additive

I think a Local trend outlier indicates a Ramp Shift, that is, the slope of the series changes at a particular point in time. However, I can't figure it out from the computational definition that SPSS provides.

I don't deal with statistics on a daily basis, and would appreciate if anyone could provide or direct me to a plain langauge description of what Innovational, Local trend, and Additive patch outliers are. Example time-series' that show what those outliers look like would be greatly helpful.

Thanks all!


No cake for spunky
This is from SPSS 17 so its possible that it changed in version 20. I am not sure it will help either :)

This section provides definitions of the outlier types used in time series modeling.

Additive. An outlier that affects a single observation. For example, a data coding
error might be identified as an additive outlier.
Level shift. An outlier that shifts all observations by a constant, starting at a
particular series point. A level shift could result from a change in policy.
Innovational. An outlier that acts as an addition to the noise term at a particular
series point. For stationary series, an innovational outlier affects several
observations. For nonstationary series, it may affect every observation starting at a
particular series point.
Transient. An outlier whose impact decays exponentially to 0.
Seasonal additive. An outlier that affects a particular observation and all
subsequent observations separated from it by one or more seasonal periods. All
such observations are affected equally. A seasonal additive outlier might occur if,
beginning in a certain year, sales are higher every January.
Local trend. An outlier that starts a local trend at a particular series point.
Additive patch. A group of two or more consecutive additive outliers. Selecting this
outlier type results in the detection of individual additive outliers in addition to
patches of them.
If it does not you might look in the following link to see if that helps. Forecasting 17.0.pdf

That does help a bit, but I'm still having trouble visualizing what an Innovational and a local trend outlier would look like.

Does anybody know, would a local trend outlier look like a Ramp shift change (that is, the slope of the series changes at a particular point in time)? Does local trend really mean a change in slope?