Dear Jens and Lars!

I just have tried to evaluate the registration ability of local seismic 
stations. And I faced with strange issue when I tried to re-calc Amp (WA 
amplitude) using hypocenter out: hdist and ml. Of course Amp was measured but I 
did it for test my scripts. 

So, what's happened:
If I re-cacl (inversely) the Amp values using Local mag equation, I never got 
original (measured) values. Just look at example (I attached hyp.out, 
hypmag.out):

 TUN  SZ   hdist:      52.3   amp:         514.5 T:   0.3   ml =   3.0
                                           562.89 <—re-calculated value

 CHIVOSZ   hdist:      67.1   amp:         211.3 T:   0.3   ml =   2.8 + -0.10
                                           216.28 <—re-calculated value

 VAL  SZ   hdist:      78.0   amp:         201.9 T:   0.3   ml =   2.9
                                           200.79 <—re-calculated value 

 NGLK SZ   hdist:     139.0   amp:          57.8 T:   0.5   ml =   2.9 +  0.06
                                            59.42 <—re-calculated value 

 NYSH SZ   hdist:     161.9   amp:          52.7 T:   0.4   ml =   3.0
                                            53.32 <—re-calculated value 

 LNSK SZ   hdist:     183.8   amp:          25.1 T:   0.4   ml =   2.8
                                            25.20 <—re-calculated value

The error always ranges from >1% to 10% according original values (+- ~0.04 
Ml). I tried to re-calc hdist using dist and depth values from hyp.out but the 
same issue. I use R but it repeated even if I use embedded OS calculator.

Why it happens? 

Best wishes,
Andrey Stepnov


STATION0.hyp:
...
RESET TEST(75)=1.00
RESET TEST(76)=1.84
RESET TEST(77)=0.0011
RESET TEST(78)=-2.97
...

/seismo/seisan/LIB/hyposub3.for 298:302

cccccccc
      magnitude=am*alog10(amp)+bm*alog10(hdist)+cm*hdist+dm
             if(mag_cor(2,kstat).ne.99.0) then
                XMAG=XMAG+MAGNITUDE + mag_cor(2,kstat)
                nml=nml+1
             endif
cccccccc


The R-function:

########

  wamp <- function(a, b, c, dist, ml) {
    LGA <- ml - (a*log10(dist) + b*dist +c)
    return(10^LGA)
  }

########




Attachment: hypmag.out
Description: Binary data

Attachment: hyp.out
Description: Binary data

_______________________________________________
seisan mailing list
[email protected]
http://mailman.uib.no/listinfo/seisan

Reply via email to