As an earthquake, and the various aftershocks, are transient events,
toggle quoted messageShow quoted text
it should be practical to take some data before and after the event
to determine the incremental phase shift to correct your long term data.
That is assuming that there has not been a jump in the value of gravity,
which should show up in the rate graph.
This is a very interesting subject,
On 24 Jan 2020, at 06:07, neil <firstname.lastname@example.org> wrote:
In New Zealand a real issue for me is the number of earthquakes we get, which really destroy long term rate tests.
On Fri, 24 Jan 2020, 4:45 AM Ian Richardson via Groups.Io, <email@example.com> wrote:
...and the Moon. When I had my most accurate clock running, the Moon's phases clearly were reflected in gaining and losing rates.
From: Johannes <firstname.lastname@example.org>
To: synchronome1 <email@example.com>
Sent: Thu, 23 Jan 2020 16:39
Subject: Re: [synchronomeelectricclock] Reinventing a Synchronome
What about what happens under the earth surface? It is always in traveling.
On 22 Jan 2020, at 12:46, Harvey Moseley <firstname.lastname@example.org> wrote:
Very nice. Beyond air density and temperature, are there any other parameters that should affect the pendulum period? Will be interesting to see the residuals after the pressure and temperature dependence are removed. Those residuals are the ones we can ascribe to the clock itself - we can't hold it accountable for the external parameters (though a good device should seek to compensate them).
On Wed, Jan 22, 2020 at 7:42 AM John Haine <email@example.com> wrote:
[Edited Message Follows]
I have obtained pressure data taken at 30s intervals at the Trinity College Clock in Cambridge (thanks to help from Dr. Hugh Hunt) over the past 11 days and got this curve of period & pressure vs. time. There is clearly a correlation though lots of other wiggles, some with a clear diurnal variation because of room temperature changes. Anyway doing a linear regression over the period shown gives a period sensitivity of 0.25 microseconds/millibar. Given the confounders that's probably roughly in line with you measurement Bepi? If I had room temperature measurements as well I could correct the density and maybe do a multivariate analysis...
Also to note that some of the variation especially in the early period is probably because of the clock settling down.