In seismology, is there a concept of a 'representative station', a
tation that collects very similar information to others? Or is the
nformation from *every* station just as valuable (informative) as all
he others?
Are there lists of representative stations?
Hi Dan,
    There are strong motion, intermediate motion and weak motion sensors
all used for different jobs. Check out www.guralp.com The usual sensors=20
detect from 50 or 30 Hz to periods of 120 seconds, exceptionally to 360=20
or 1000 seconds. If you click on stations shown on a world map, say USGS
or Iris, this calls up details of the seismometers being used at that=20
station.
If I were the funding god, and I told you to axe all but 20, 50, 200,
r 500 stations, which stations would you choose to keep and why?
 =20
    You want to be able to see quite small quakes as well as
the large ones and the signal decreases in amplitude as the=20
distance increases. There are large areas in the oceans where you=20
can't installqa seismometer. Cost is the other limitation. the=20
instruments are expensive, they have high installation costs, they=20
need to be powered and the signals need to be relayed and monitored.

y thinking is that 10 stations in one relatively small area must be
ollecting similar data, and you can (perhaps) represent *nearly* all
0 by picking just one of them (the most 'representative' one). Each
new' station must add more information, but at some point, the amount
f information gain from each additional station must plateau.
Does any of that make any sense?
    Not a whole lot.=20
Regards,
Chris Chspman  =20







  In seismology, is there a concept o=
f a 'representative station', a
station that collects very similar information to oth=
ers? Or is the
information from *every* station just as valuable (in=
formative) as all
the others?

Are there lists of representative stat=
ions?
Hi =
Dan,
    =
There are strong motion, intermediate motion and weak motion sensors=
all used for di=
fferent jobs. Check out www.guralp.com The usual sensors 
detect from 50 or 30 Hz to periods of 120 seconds, exceptionall=
y to 360 
or 1000 seconds. If you click on stations shown on a world map, say USGS
or Iris, t=
his calls up details of the seismometers being used at that 
station.

If I were the funding god, and I told you to axe all =
but 20, 50, 200,
or 500 stations, which stations would you choose to k=
eep and why?
   
    You want to be able to see quite small =
quakes as well as
the large ones and the signal decreases in amplitude as the 
distance incre=
ases. There are large areas in the oceans where you 
can't installqa seismometer. Co=
st is the other limitation. the 
instruments are expensive, they have high installat=
ion costs, they 
need to be powered and the signals need to be relayed and monitored=
..

My thinking is that 10 stations=
 in one relatively small area must be
collecting similar data, and you can (perhaps) represent *nearly* all
10 by picking just one of them (the most 'representat=
ive' one). Each
'new' station must add more information, but at some =
point, the amount
of information gain from each additional station must=
 plateau.

Does any of that make any sense?

    Not a whole lot. 
Regards,
Chris Chspman