GPS Tracking

I recently downloaded an app on my Android phone that logs GPS coordinates as I’m traveling. I was a bit unsure how well it would work, but a few test drives proved that it’s quite capable, at least in areas where cell phone coverage is good. I live in the Phoenix, Arizona area, and go hiking in South Mountain Park quite often. Since the park is just south of the metro area, and one of the mountains is peppered with cell phone towers, coverage is very good there.

So I gave the phone a try, tracking my progress around a two and a half mile hike over relatively easy terrain. I then exported the data from the web-site where it is stored, converted it to KML format so that Google Earth could read it in, and plotted my path. The screenshot below shows the track I took, the elevation that was traversed, and has a number of hot-spots where I stopped to take a few photos. The latter elements were added manually, which was tedious, but for me at least, makes the plot more interesting.

I’ve tried a tracking myself a few times since the hike, and am fascinated by the possibilities. I will continue to track my hikes this spring, and I think I’ll also try tracking commute time and speed, both of which are also recorded. I’ve tried doing this in the past manually, but that often meant that I forgot to check my watch, resulting in very incomplete data. Using the phone, all I need to do is remember to turn tracking on; should be pretty easy.

As an alternative, I’ve added an interactive Google Earth interface below. You can control the view in the same manner as the Google Earth application.


With the new Kepler telescope findings having hit the news-feeds this month, I thought I would post an update to my original data. The tally of Kepler planetary candidates now stands at 2,326, with one particularly noteworthy find: Kepler-22b. Also known as Kepler Object of Interest (KOI) 87.01, or Kepler Input Catalog ID 10593626, this planet deserves special attention, as it is one of the first confirmed planets that resides within the habitable zone of its star. At only 2.4 times Earth’s diameter, and with an orbital period of around 290 days, it seems to be a pretty close match to the Earth.

The Kepler web-site has a nice article describing the planet, with the image below showing how it stacks up against the inner planets in our own solar system.

NASA Mission Page – Kepler (image credit: NASA/Ames/JPL-Caltech)

After reading about the new planet, and trying to find the raw data for the most recent 1,000+ planetary candidates to no avail, I got to thinking about the process that is used to detect transits like those for Kepler-22b. I know it’s pretty tricky, not just because of the miniscule amount of dimming due to the planet passing in front of the star, but also because the stars themselves have turned out to be more variable than anyone suspected. In addition, there are changes in the array of detectors on the spacecraft that must be nulled out, and a host of other considerations.

Given all that, I pulled the corrected light curves from Kepler-22b from the NASA site where the data is stored. The plot below shows the results. Looking at this data, which has already gone through a significant transformation process, should give you some idea of the difficulty in finding small planets.

Raw, Corrected Light Curves – NASA MAST web-site

Finally, if we normalize each successive observation to the median for that run, this allows us to plot all the data on one scale in order to compare each and look for the tell-tale dips that signal a planetary transit. It’s surely not intuitively obvious when the transits occurred, so I’ve pointed them out with arrows. The smaller, embedded plot is a zoom in on each of the transits, showing how similar they appear. This fact, coupled with the a common interval of 289.9 days between them, gives scientists high confidence they’ve found their target.

Normalized Light Curves – Data processed with Matlab R2011b

There are many articles on the web that describe Kepler-22b and the remaining planetary candidates. One of my favorites is from the Sky & Telescope web-site. And all the data that’s available on the web makes it easy to forget that there’s an actual star out there that is being measured. The scientific journal article had a nice picture of the star, Kepler-22, reproduced below.

“Kepler-22b: A 2.4 Earth-radius Planet in the Habitable Zone of a Sun-like Star”; Borucki, et al; eprint arXiv:1112.1640

Earthquake Analytics

I’ve been experimenting with various ways to present data in this blog, mainly using static screen shots of graphs. But over the weekend, I started exploring interactive options. In particular, there is a very capable tool out there called Tableau that provides a way to publish analytic graphics to the web, using their publicly available server to process the data. It has quite a lot of flexibility. The analyses below are my first foray into this endeavor. I also updated my earlier post on the Kepler telescope results to make it more interactive as well.

All data is from the U.S. Geological Survey earthquake database.

This first visualization shows the world-wide distribution of earthquakes from January 1, 1973 to October 14, 2011 from magnitude 4.9 to the maximum recorded of 9.1. The tectonic plate boundaries are easily visible. You can change the magnitude and year sliders to limit the data to ranges of values as desired. When you hover over the map, additional controls are available that will allow you to zoom in on specific regions of the map.

The second visualization shows the most recent few years of earthquake activity around Japan. Scroll through each year by clicking the arrow button under the control labeled “Recent”. Note the significant increase in earthquake frequencies in 2011, centered off the coast of Tohoku, the site of the devastating earthquake that struck on March 11.