Earth-prints repository, logo   DSpace

About DSpace Software
|earth-prints home page | roma library | bologna library | catania library | milano library | napoli library | palermo library
Please use this identifier to cite or link to this item: http://hdl.handle.net/2122/7515

Authors: Bindi, D.*
Pacor, F.*
Luzi, L.*
Puglia, R.*
Massa, M.*
Ameri, G.*
Paolucci, R.*
Title: Ground Motion Prediction Equations Derived from the Italian Strong Motion Database
Title of journal: Bullettin Earthquake Engineering
Series/Report no.: 6/9 (2011)
Publisher: Elsevier
Issue Date: Dec-2011
DOI: 10.1007/s10518-011-9313-z
Keywords: Ground motion Prediction Equation
ITACA
Abstract: We present a set of ground motion prediction equations (GMPEs) derived for the geometrical mean of the horizontal components and the vertical, considering the latest release of the strong motion database for Italy. The regressions are performed over the magnitude range 4–6.9 and considering distances up to 200 km. The equations are derived for peak ground acceleration (PGA), peak ground velocity (PGV) and 5%-damped spectral acceleration at periods between 0.04 and 2 s. The total standard deviation (sigma) varies between 0.34 and 0.38 log10 unit, confirming the large variability of ground shaking parameters when regional data sets containing small to moderate magnitude events (M < 6) are used. The between-stations variability provides the largest values for periods shorter than 0.2 s while, for longer periods, the between-events and between-stations distributions of error provide similar contribution to the total variability
Appears in Collections:04.06.10. Instruments and techniques
Papers Published / Papers in press

Files in This Item:

File SizeFormatVisibility
Bindi_ITACA_GMPE.pdf1.51 MBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.


Share this record
Del.icio.us

Citeulike

Connotea

Facebook

Stumble it!

reddit


 

Valid XHTML 1.0! ICT Support, development & maintenance are provided by CINECA. Powered on DSpace Software. CINECA