Reducing the subjectivity in Modified Mercalli intensity analysis

Reducing the subjectivity in Modified Mercalli intensity analysis: drawing isoseismals Warwick Smith

MM intensities are arguably: • • • Subjective Useless Hopeless Pointless Meaningless All of the above But they are all we have for our most important earthquakes, and ever will have. We had better learn all we can from them. NZSEE Annual Conference 2007 GNS Science

Hawke’s Bay 1931 G. A. Eiby NZSEE Annual Conference 2007 GNS Science

Why the Roman numerals? Maybe …… 1. To distinguish intensities from magnitudes 2. To discourage fractions – use integers only What is 6. 7 in Roman numerals? VI VII X My practice: Arabic for individual observations Roman for isoseismals NZSEE Annual Conference 2007 GNS Science

Hawke’s Bay 1931 G. A. Eiby NZSEE Annual Conference 2007 GNS Science

Drawing isoseismals The developing practice is to model them as ellipses in an attenuation function Let’s try to fit them as ellipses NZSEE Annual Conference 2007 GNS Science

Which is the best ellipse for MM VI? Minimise the sum of distances to anomalous observations NZSEE Annual Conference 2007 GNS Science

Assigning weights for fitting MM VI isoseismal Intensity Weight if inside Weight if outside 4 2 0 5 1 0 6 0 1 7 0 2 8 0 3 NZSEE Annual Conference 2007 GNS Science

Then it gets more complicated Intensity Weight if inside Weight if outside 4 -5 1. 5 0 5 -6 0. 5 6 -7 0 1. 5 7 -8 0 2. 5 NZSEE Annual Conference 2007 GNS Science

Then it gets more complicated Intensity Weight if inside Weight if outside 4 -5 1. 5 0 5 -6 0. 5 6 -7 0 1. 5 7 -8 0 2. 5 4? 1. 5 0 5? 0. 5 0 6? 0 0. 5 7? 0 1. 5 8? 0 2. 5 NZSEE Annual Conference 2007 GNS Science

Parameters to fit • Coordinates of centre (lat, long) • Axes (major, minor) • Orientation But for a stable solution we can specify the centre and require that all the isoseismals have the same orientation. NZSEE Annual Conference 2007 GNS Science

Hawke’s Bay 1931 NZSEE Annual Conference 2007 GNS Science

Buller 1929 Very poor control NZSEE Annual Conference 2007 GNS Science

Why are they not circles? • Long fault sources? • Anisotropic propagation? NZSEE Annual Conference 2007 GNS Science

Why are they not circles? • Long fault sources? • Anisotropic propagation Energy propagation parallel to the strike of the country seems to be more favourable than across it. Source effects should mostly be seen at short distances, comparable with the fault length. NZSEE Annual Conference 2007 GNS Science

Hope 1888 NZSEE Annual Conference 2007 GNS Science

Orientation of major axis Rather than solving for the strike, should we use: • A representative regional value for the outer isoseismals • The fault strike for inner isoseismals • Interpolation between ? ? ? ? Maybe NZSEE Annual Conference 2007 ? GNS Science

Systematic analysis All large events (e. g. Dowrick & Rhoades) Digitize all reporting locations, with intensities Fit isoseismals to each Watch this space Fit attenuation model to derived semi-axes NZSEE Annual Conference 2007 GNS Science

What happens when ellipses don’t fit well? • Soft ground gives systematically high intensities (e. g. Wanganui) • Strong attenuation gives low intensities (e. g. across the TVZ) • etc Procedure • Fit ellipses • Look for systematic departures NZSEE Annual Conference 2007 GNS Science