I use the description ``regional-distance'' to denote data that are
recorded at distances greater than the near-source region and less than
30. There has been a great deal of research in this area that has been
motivated by the deployment of broadband digital networks over wide
areas of the United States and other regions. Seismograms recorded at
regional distances are complicated by strong interaction with the
heterogeneous crust and upper mantle. Thus, calculating accurate Green's
functions, and hence modeling seismic sources using these data, is more
challenging than at greater or lesser distances. Moment-tensor inversion
methods, which have met with great success in studying teleseismic
events, were applied to these data with the result that the threshold for
moment-tensor determination has been lowered from the 
5.5 level for teleseismic events to approximately 
4.0
level for regional events. Frequency-magnitude relations lead to the
obvious conclusion that this will allow seismologists to increase the
number of events studied by more than an order of magnitude in areas with
regional broadband instrumentation. The signal-to-noise ratio at long
periods required for such modeling will make it increasingly difficult
to push this limit lower. Thus, for micro-earthquakes (
3.0) seismologists are likely to have to continue to rely on first-motion
focal mechanisms from dense, high-gain networks.
Romanowicz et al. [1993] modeled the spectra of regional
surface waves in a study of earthquakes recorded in central and northern
California (
3.4). Patton and Zandt [1991]
used a similar approach to study regional surface waves in the period range
7-40s and determined focal mechanisms of events in the western United States
in the local magnitude range (3.5 
5.8).
Thio and Kanamori [1992] used surface waves recorded at
regional distances in southern California to study moderate earthquakes.
By using a regionalized propagation model, they were able to study events
as small as 
=3.0 using a sparse network.
Beck and Patton [1991] also used surface wave spectra
to study aftershocks of the Loma
Prieta earthquake at regional distances. Ritsema and Lay
[1993] used a program developed by Kawakatsu [1989]
to determine moment tensors for 
5.0 earthquakes in the western
United States using long-period surface waves (T 
35 s). Although their
magnitude threshold is not much lower than that of teleseismic catalogs,
using regional data allowed them to calculate moment tensors more
rapidly. Giardini et al. [1993] determined moment tensors
for earthquakes (
5.2) in the Mediterranean region with a
regional array.
Dreger and Helmberger [1991a,b, 1993] used body waves that interact strongly with layering in the crust to model moderate earthquakes in central and southern California using data from sparse three-component broadband networks. They found that these phases provide useful information on source properties because they are relatively insensitive to short-wavelength lateral heterogeneity. In areas where the crustal velocity structure is well known, Dreger and Helmberger [1993] were able to obtain stable results using only two stations even for events outside of their array.
Permanent broadband networks and deployment of temporary broadband
networks during aftershock sequences typically have large station
spacing. During the last four years, several investigators have explored
the use of either single stations or a very small number of stations to
determine source parameters at local and regional distances.
Fan and Wallace [1991] used a moment tensor inversion
on local earthquakes as small as 
=3.5.
Fan et al. [1993]
also applied this procedure to aftershocks of the 1991 Costa Rica
earthquake. Uhrhammer [1992] and
Romanowicz et al. [1993] used near-field recordings of earthquakes at a
single station to determine the moment tensor of events as small as
=3.2. Walter [1993] used broadband three-component
recordings from a single station at a distance of approximately 300 km to
determine the source parameters of the Little Skull Mountain earthquake.
Ma and Kanamori [1991] were able to determine
fault-plane solutions for aftershocks as small as 
=1.9 located on the
Raymond Fault directly underneath the broadband station in Pasadena.