Contour units are in g and the red numbers show observed PGA in units of %g.  Red circles show where predicted PGA values are used in the contouring of the shakemap.  The predictions are made every 22 km and used only when there is not strong motion station located within 22 km.  Thus if there is a recording available it is used in the shakemap generation and the model values are used only where there are no observations.  Following the method outlined by Wald et al. (1999) we interpolate the sparse data/model grid shown prior to contouring the map.  This tends to smooth the integration of observed and predicted values.
	The top row compares (left) a data-only shakemap calculated using the Lucerne Valley recording and one without Lucerne Valley (right).  The bottom row compares data/model shakemaps with (left) and without (right) Lucerne Valley.