Jillian's blog

So I've been trying to calculate the apparent resistivity for our sites to be sure we know the electrodes are working well and so that we have a baseline to work off of when there is a fluctuation in the field. My problem is that I got this crazy graph of apparent resistivity as a function of frequency that oscillates like crazy (in more quantitative terms "like crazy" translates to about 10 orders of magnitude). I thought my problem was that I wasn't taking into account the Volts per nanotesla on our magnetometers. I spent a bunch of time re-scripting and fitting the response curve for the magnetometers. I then also found a response curve for the elctrodes and applied that in the same way. After all of this work, I'm back go a graph that looks about the same as the one I started with. The only difference is this one is translated up about 10 orders of magnitude. So now, not only to my numbers change too much at a given frequency, but they're in general too high (unless we happened to burry our electrodes in a piece of steal, but this I sincerely doubt since my adviser was there burrying them).

While I set this to rest for a bit, I can spend my time becomming frustrated with how to download several hundred datasets from Berkeley's servers without doing it manually. I'm thinking UNIX script, but its been a while since I've used one so I'll need to brush up. Plus, my work computer runs on Windows, so I'll likely have to use one of our other machines for this.

On top of all this, the magnetometer we were going to dig up and re-bury can't be because we don'thave all the necessary pieces. These pieces, of course, need to be custom fabricated by a guy at USGS who is out of town for a week. So back to the holding pattern on that part of the project.

After being away for 2 weeks and then sick for one, I'm finally back to making good progress. My very first Matlab script ever is working: not only is it running without errors, its calculating and graphing what I want it to.

I did a weekly schedule back at the beginning, but to summarize goals (per Michael's request) here they are:

first third: learn matlab, read papers, begin planning re-installation of E-W magnetometer

second third: calculate apparent resistivity for at least Jasper Ridge, Berkeley, and Marin Headlands sites, hopefully find something interesting about instrument stability, test Jarrett's new board that he's building

third third: get AGU poster ready, finish testing board and calculating resistivities

This will, of course, all change if we get a big earthquake. So everyone cross their fingers that we get a mag 6+ here in the bay! (minus any serious destruction)

Now back to Matlab...I must make my script more fancy so its everything I dreamed it to be!

Yesterday was my first day with my adviser, Simon here. He helped me get much more focused on the project and even some specific things to work on over the next few days. I had my first success at about 9:30 this morning: I was finally able to figure out how to download our data from Berkeley. (That's right Berkeley and Stanford on the same side...although maybe with some hostility since I had a lot of trouble at first). I am currently trying to expand my Matlab knowledge since I'm going to be writing a lot of scripts in Matlab this summer. I had trouble with the Fourier transforms because our data doesn't have time included in the array (like the data we worked with at orientation). Instead, the start time and frequency are in the data file name and the file only contains the actual EM data. So...FOR LOOPS to the rescue! Being the expert programmer I am, I decided a for loop would be a great way to create an array of time values based on the start time and frequency. One minor problem with this brilliant idea is that I'm working with a data file that is a whole day of data at 40 Hz. That adds up (or I guess multiplies up is more accurate) to 3,456,000 data points which means my for loop is trying to create that many time values in an array. It’s been running now for about 2 hours and I'm still waiting. Unfortunately I'm running out of other things to do without Matlab (hence the blog entry).