I have noticed a few times that the Big Red Button on the ISI_CONFIG screen is a little too appealing, and the button has been used at times when it maybe shouldn't have. The original intent of that button was to put the platforms in the safest possible state before or during a Montana sized (as in the 5.8 magnitude in Montana during July of O2) earthquake. The earthquake last night tripped ISI's, but probably wouldn't have tripped HEPI. So, I've changed the script the button runs to just take the platforms down to damped, leaving HEPI on. This should save some recovery time in the future. 

Because seismon doesn't give a good prediction of what surface velocities we should expect, the verbal notification only gets triggered if the earthquake is above a 6. So what should we do when we get a verbal notification? It depends, every earthquake is unique, but generally the smaller and further away the earthquake is, the response should be less dramatic. I discuss a few cases here.

1. Really, really bad earthquakes like the 5.8 in Montana, very close. Something like that will be onsite before seismon will see it.

Response: do nothing, wait for things to settle down before trying to recover.

2.Really bad farther away like 7-8+ in Mexico or Alaska, large and on the order of 5-6000 km away. 

Response: Put the platforms in as safe a state as possible, we will probably get a couple minutes early warning,. The Big Red Button will start this, but if the earthquake is big enough, or close enough, the added step of turning HEPI off may be help. But if the ISIs are already damped, it may not be any worse to just let HEPI trip on its own.

3. The 7 in Peru last night is kind of an edge case, about 8000 km away.

Response: Probably put the SEI_CONF guardian in the LARGE_EQ_NOBRSXY state. This won't save a lock, but this should keep ISI's from tripping, especially if the earthquake is much farther away than 8000 km. The Peru earthquake tripped ISIs, but I can't say for sure that changing just the sensor correction state wouldn't have prevented that. But, being on a different continental plate reduces the effect on we see here, I believe similar sized earthquake a bit further north in Central America would have been worse for us.

4. Magnitude 6-7 more than 8000 km away. These are common, sometimes we ride them out and they don't present any risk to the IFO.

Response: Nothing, or try the SEI_CONF EARTH_QUAKE state. Again, smaller and farther away require less response. We generally don't even see 4-5 magnitude eqs over 8000 km away.

I haven't had many chances to test it, but there is some fancy new code in the seismic models that we hope will allow us to stay locked through earthquakes like this. Because most of the ground motion is common ground motion during an earthquake is common mode, the new code uses the corner and end station seismometers to calculate this common motion, and the models subtract this out of the sensor correction signal. In SEI_CONF, this should be the EARTH_QUAKE state. I haven't had a chance to test this during any earthquakes, and it will probably help only during moderate or low microseism times.