Homestake Meeting, 01/09/2014
Attending: Victor, Vuk, Gwynne, Tanner, Gary, Gabor S, Shivaraj
1) PASSCAL communications (Vuk, Gary, Tanner) Gary spoke with Tim Parker at PASSCAL. Several outcomes:
a) We have all the wireless radios we want to use. They have more than the number we need for surface sites. Have anthena on Yates, and line of sight to the surface stations. We need to discuss if they could play a role in some underground sites. That is, if a door isn't it the way we could potentially deploy instruments totally autonomously and do comms wireless down the drift. Vuk: attempted wireless communication in the drifts in the early days, didn't work well. Also, there are usually turns in the drifts, difficult to get a direct line of sight.
b) Multiple solutions exist for sending time that are off the shelf. We need to just get started on that one. Vuk: this is the highest priority now. There are solutions developed by other users, but may require us to develop some hardware. There may also be vendors selling repeaters of the GPS signals, eg used for subway tunnels etc. We would need to speak with the vendors to understand exactly what signals are being repeated, and make sure this is what the Q330 would require...
c) Disposable batteries: air cells are an option to consider. Rough number I got from Tim Parker was $100/month for the station alone and double that if we have to feed a radio or a fiber optic device (one or the other or both are required of course underground). Lighter than standard acid batteries, about 1/4 weight, and disposable. Vuk: batteries may really be the way to go. All of the locations we are considering are accessible via rail or other vehicles. So it shouldn't be difficult to replace ~10 batteries in a station every ~6 months or so.
Gary: Also, maybe could use DC voltage supply over a ~2km cable. Should check with SURF whether there are cost savings if we go with this option, as opposed to AC power. Shivaraj: there are constraints due to fire-hazard on the insulation material etc, which drive the costs up.
2) UMN test setup (Tanner) Tanner: Received the first Q330 from PASSCAL, starting to set it up in the lab at Minnesota. Should have more to say next meeting. Vuk: the idea is to gain experience with this system and learn how to use the GPS with Q330. Then we can start talking to vendors and others to understand how to set up synchronization.
3) Analysis of existing data
a) Frames/miniSEED files (Shivaraj, Tanner, Michael) Shivaraj: produced frames for another 2 months of data, currently at Caltech. Continuing to produce frames for other data. Timing synchronization is at 1 ms, so not too bad. But data usually does not include all stations. Also, observe DC level shifting over time in discontinuous steps, on varying time-scales (sometimes even shorter than 10 sec). Gary: are you dropping bits? Are the steps powers of two? Are there ripples in the series? Shivaraj: appear to be sharp jumps. Gary: so it is probably digital, otherwise there would be some ringing in the time series. Check if these are power of 2 jumps... Shivaraj: see it in 4100A, could be present in other stations as well. Suspect we should be able to find another good patch of data.
b) Wiener filtering (Michael, Jan)
c) Estimation algorithms (Shivaraj, Eric) Shivaraj: planning to add P and Rayleigh filters into 1 filter. Vuk: Had a discussion with Eric and Shivaraj about this. Roughly, Shivaraj will aim to complete his study by combining the P and R filters into one, test it on simulations, and then apply it to the real data acquired at Homestake. This should be sufficient for a publication. Eric will take a longer-term approach, trying to incorporate all wave modes, address computational issues that will probably arise, and study possible degeneracies between different modes.