Wednesday, 8 September 2010

Canterbury Earthquakes part III

As more people keep following the GeoNet website, they are starting to notice the seismic activity in areas other than Christchurch. A few people I've talked to have expressed concern that the Saturday 7-pointer has sparked earthquakes around the country. This is incorrect, as earthquakes below magnitude 3 are extremely common and as can be seen in the plot below, they occur throughout the country.

A closer look at the magnitude of these earthquakes shows their temporal distribution is fairly uniform. What is interesting is that prior to the time of the big one there appears to be a lull in the frequency of lower-level earthquakes nation-wide. Is this at all significant? I don't know.

The magnitude follow a right-skewed normal distribution with a mean of 2.47 and a standard deviation of 0.617, while their depth follows an approximate Poisson distribution with a mean of 44 km.

Looking at the seismic activity within the Canterbury region, we see that earthquakes appear to have been fairly randomly distributed throughout the region until the 4 September quake. Only a single tremor was detected from the vicinity of the recently revealed fault, an insignificant 2.31 that occurred on the 21st of March. That there weren't more tremors closer to the time would seem to count against the hypothesis of big earthquakes being preceeded by smaller ones.

As before, data was gained from the Quake Search data download query provided by GNS, and the data, R code and a file giving the location of Canterbury towns is available for all.


Efrique said...

... a right-skewed normal distribution

Since the normal distribution is symmetric, being right-skewed means it's not normal; you don't get to have both.

nzcoops said...

It's just a description. Think right-skewed bell shaped distribution. Couldn't have 'right skewed distribution', has to be another descriptor in there.

Samuel Brown said...

Cheers guys. Absolutely correct. Should've been a bit more sharp on the language!

Jesmin said...

bioinformatics training chandigarh

I also study statistics and compltetely agree with Efrique...please explain the difference