Homestake Telecon Minutes, 10/17/13
Attending: Vuk, Eric, Riccardo, Tanner, Pat, Jan, Nelson, Victor, Michael, Shivaraj
1) Status at Homestake (Vuk, Victor, Tanner) Vuk: visited Homestake with Victor and Tanner in mid Sep. Extracted and returned two seismometers to NIKHEF and one to Minnesota (along with some electronics). Also scouted a number of locations at 1700, 2000, 4100 and 4850 levels. Found that many of the locations would be good for seismic stations, and are currently in the process of estimating costs for them. Once the cost estimates are ready, we will determine which ones we want to develop, optimizing the science while fitting within the budget. The locations come with very variable costs, depending on the amount of ground support that is needed, and on the distance to the nearest power and network.
2) PASSCAL communications (Vuk, Gary)
a) availability of instruments b) timing/synchronization
Vuk: with Gary, we had several contacts with the PASSCAL representatives. They will be able to provide seismometers in Dec 2014, and we will be able to keep them for 2 years. We should be able to get some digitizers much earlier, which will be necessary to develop the timing/synchronization scheme. The digitizers are typically relying on the GPS signals for timing, which we don't have easy access to. This problem has been solved in the past by other projects. PASSCAL people are currently trying to identify which digitizer technology would be best for this, and will try to provide us with information about how the problem was solved in the past. We hope to get one or two digitizers soon, so that we can start looking into this problem.
3) UMN test setup (Tanner) Tanner: have one seismometer from Homestake, and much of the accompanying electronics. Currently trying to put the system together and start playing with the existing data acquisition (with Shivaraj's help). The idea is to develop a prototype at UMN of the seismometer+DAQ system, including the solution to the timing problem, so that we can prepare a solution for Homestake before the seismometers become available.
4) Analysis of existing data
a) Frames/miniSEED files (Shivaraj, Tanner, Michael)
Shivaraj/Vuk: Two years of data exists, with gaps, probably about 1 year of data integrated. In ASCII it is about 1TB, but should be significantly less in frame and miniSEED formats. Shivaraj currently in the process of producing frame format files. Have data for one good month (Feb 2012) already available.
Michael produced the miniSEED files for the entire month, should be available at Caltech. At UMN we have only a few selected miniSEED files, which Gary could access fine. Gary brought up an issue with channel naming, Michael will look into it and find the best way to deal with it. Then we should find the best way for Gary and Victor to access the miniSEED data.
b) Wiener filtering (Michael, Jan)
Michael: lately looking at LIGO data, would like to have tilt-free estimate of the horizontal component. There are ways to do this using the vertical channel, using Hilbert transform. Trying to implement it now. Looking at this as another alternative to the Wiener filtering, which we know works.
Jan: Assuming that you have a Rayleigh wave and that you know the amplitude, you can estimate the tilt effect in the seismometer. The horizontal and vertical components of the R-wave are related via the Hilbert transform, which is acausal. So need an causal filter approximation which may or may not work.
Michael will provide more detail next time.
Jan: regarding the paper, we have shown that we can have significant suppression with the Wiener filter. But there are interesting questions that one could ask about the microseisms: could study the variation of the filter coefficients as a function of time, eg if you update it every day for a month. Check if the coefficients adjust to the changes in microseism.
Jan: probably should split the two papers, but we can wait with the decision. Vuk: agrees.
c) Estimation algorithms (Shivaraj)
Shivaraj: agrees that we should split the two papers. Short term, the focus is on the P- and R-waves, will try to implement a joint filter in the radiometer code and try it on real data. But there are other aspects, such as including the S-waves in the simulation etc. Probably could benefit from additional manpower.
Will put the code in a repository, may be in matapps already.
5) New wiki (Vuk) Will start a new wiki, supported at UMN, where we can keep the relevant materials (meeting minutes, presentations, papers, information about stations etc). Will require username and password for each person who wants access to the wiki. Vuk will provide the names and email addresses to the system administrators, and they will provide access.