[Shake-dev] amplification on soil sites / MMI bias

[Bruce Worden]

Hi Georgia,

It's a little hard to tell what's going on, but it looks to me like  
much of the blue spottiness is associated with the stations on the  
map. Looking at the plotregr plot, it appears that many of the  
stations (even at distance) are intensity > 1.0. Even past 200 km,  
some appear to be as high as 1.3 or so. These will not be white on the  
intensity map, and nearby site amplification could bring them 0.3  
intensity units higher. I am a bit surprised that their influence is  
as great as it is, however. What GMICE and IPE are you using? What  
does them map look like if you turn off site amplification?

I think bias is probably not particularly accurate for these very  
small events, but if you use it, you should probably restrict the  
maximum radius to a much smaller distance. It is normally set to  
something like the accurate range of your GMPE, but for small  
magnitudes, the PGMs are probably near the noise level much closer to  
the epicenter than that.

Bruce

On Feb 22, 2011, at 5:37 AM, Georgia Cua wrote:

> Dear Bruce,
>
> We are now running grind (id 313 2011-01-24).
>
> We've rerun the makeMMItables script.
>
> However, we still have "patchy" intensity maps that look like this.  
> (The "patches" are less pronounced than produced in the previous  
> version of grind, but there are still blue patches at relatively  
> large distances from the epicenter.)
>
> It is still probably that effect that you mentioned before on  
> applying site amplification on intensities?
>
> Is this what would be expected with the most recent fix, or did you  
> expect that the far-away blue patches would go away completely?
>
> Just for reference, I'm also attaching the mi_regr plot. The  
> intensity bias is -0.83. It is in the right direction now, with the  
> recent fix. Does a bias calculation make sense when the data are  
> essentially on a flat line? (One could still argue for no bias  
> calculation in this case?)
>
> Thanks, and sorry to keep bothering you with these not-felt/weak  
> shaking events!
> Georgia
> <intensity.jpg>
> <mi_regr.png>
>
> -----------------------------
> Dr. Georgia Cua
> Swiss Seismological Service
> Institute of Geophysics
> ETH Zurich, NO H61
> Sonneggstr 5
> CH-8092 Zurich
> Switzerland
>
> Tel:  +41-44-633-7574
> Fax: +41-44-633-1065
>
> georgia.cua at sed.ethz.ch
> -----------------------------
>
>
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://geohazards.usgs.gov/pipermail/shake-dev/attachments/20110224/1a449b28/attachment.html>