Tuesday, May 8, 2018

Comparing two 6-hour periods of tremor: Same seismograph


Now that you have your ImageJ software installed, answer this question:
On Kilauea Volcano there is a seismograph near a geothermal plant at risk.
For May 22, and 23 2018 were the tremors increasing or decreasing?  Use the two 6-hour snapshots provided.

Here are the two seismograph reports to look at:



May 22nd, 2018

This is a very simple question to answer since it only looks at relative tremor amounts. There is no need to subtract background noise. Further since an entire six hour block of data is being looked at there are only two data points.

                 Area        Mean
May 22 1090800 71.618 0 255 0 255
May 23 1090800 71.514 0 255 0 255

Answer to the question is decreasing. Total time to answer the question was 3 minutes including loading the software.

The mean value provided by ImageJ indicates a downward trend.

What about the pixels associated with the page itself?

Answer: In both cases all the background pixels associated with the lines and numbers are also being counted. In the previous blog we discussed finding that the mean for the entire page is 4.859.

Lets convert that to a percentage of pixels: 4.859/255*100=1.905%

While we are at it lets convert mean values for May 22-23, 2018 to percentages too:

May 22, 2018: 71.618/255*100=28.085%
May 23, 2018: 71.514/255*100=28.045%

If we subtract out the pixels in the page we get:

May 22, 2018: = 26.18%
May 23, 2018: = 26.14%

What about the normal background noise and line?
Here is where things get a little difficult. We have no baseline of an entire page with "normal" background noise. What we do have to work with is just a 15 minute window on a seismograph located on Mauna Kea. That value found in the previous page was:

  • Mean 32.27 for a 15 minute segment with an area of 34500 pixels
Lets figure out how many pixels that is:

Answer: 32.27/255*34500= 4,366 Pixels

During a 6 hour period, there are 6*4=24 lines across the page.

Therefore, 24 lines * 4,366 Pixels = 104,784 pixels

Note: If you are really paying attention, you know some of those 104,784 pixels are associated with the blank page. Actually, I already helped you find out how many if you look back at that page. Correcting for the background pixels already subtracted we get a corrected value of 92,664*.

Now lets calculate the net number of black pixels associated with tremor.

May 22, 2018: (1090800 pixels *  26.18%) - 92,664 =  192,907 pixels
May 23, 2018: (1090800 pixels *  26.14%) - 92,664 =  192,471 pixels

Therefore, after subtraction of background noise we arrive at corrected difference between the two dates.

The percentage decrease was approximately (192,907-192,471)/192,907*100%=0.23%

*92,664 Pixels for a six hour page with normal background noise will be used going forward to help calculate the net tremor for a seismograph. This value will be subject to correction going forward as better approximations of background noise are found.

No comments:

Post a Comment

I greatly appreciate comments and will read all. However, I do moderate content and have to check what you write before it is posted live.

Comparing siesmic tremor with surface events

Three events late May 2018 were added to the graph of seismic data. Time for this graph was adjusted so data points represent the end of a 6...