Slashdot Mirror

← Back to Stories (view on slashdot.org)

Local Atmosphere Heated Rapidly Before Japan Quake

Posted by Soulskill on from the shake-and-bake dept.

eldavojohn writes "A new paper presented at NASA's Goddard Space Flight Center in Maryland shows the rapid heating of the atmosphere directly above the fault days before the devastating earthquake hit. This is theorized to be the Lithosphere-Atmosphere-Ionosphere Coupling mechanism that occurs when large amounts of radon are released due to massive stress in the fault right before the quake. This can be detected with satellites analyzing infrared waves: 'The radioactivity from this gas ionizes the air on a large scale and this has a number of knock on effects. Since water molecules are attracted to ions in the air, ionization triggers the large scale condensation of water. But the process of condensation also releases heat and it is this that causes infrared emissions.' This is a shift from the Haiti earthquake where DEMETER was used to monitor ultra low frequencies. The presence of radon could also possibly explain erratic wildlife behavior prior to an earthquake."

3 of 202 comments (clear)

  1. it traveled for miles through the water? by gumbi+west · · Score: 5, Insightful

    Radon is a gas and that part of the ocean is very deep. How would it have traveled a few miles to the surface so quickly and without dissolving? You might think that all noble gases are not soluble in water, but radon is actually fairly soluble.

  2. No it did not. by goodmanj · · Score: 5, Interesting

    No, the atmosphere did not heat up rapidly as a result of the quake. This article is total bullshit.

    1) Geology: There is no "buildup of unusual stresses" in the days before an earthquake. The stresses build up over decades: the only thing that changes suddenly is the Earth's motion in response to them.

    2) Oceanography: Any radioactive gases released by the fault (the mechanism claimed by the authors) would be released *at the bottom of the ocean*. From there it would have to dissolve in the ocean and be carried to the surface. This takes a *LONG* time.

    3) Meteorology: Any gases released will mix rapidly in the atmosphere, forming a plume stretching hundreds of miles from the source in a matter of hours. It will not form a coherent blob hovering over the fault.

    4) Statistics : the plot in question is supposedly based on "NOAA OLR data". It's been massaged to within an inch of its life, using a statistical technique which is highly sensitive to what happened not just during 2011, but to the vagaries of weather in 2006-2010. The result is a massive exercise in small-number statistics, which is then amplified by:

    5) Data visualization: Notice that the OLR "spikes" form nice concentric circles, and they seem to line up along a latitude line. Why? Because what you're seeing is data smoothed to a radius smaller than the actual size of the atmosphere being measured. The link below is to the *actual* raw NOAA AVHRR OLR data over Japan: there are only 9 real data points in the field of view shown by TFA, and they do not show any sign of a peak in OLR over northern Japan.

    https://picasaweb.google.com/lh/photo/veC_EraWL5NUXaCbH6iROcyKBwp3MOnR9qYUE-fJ7v0?feat=directlink

    1. Re:No it did not. by goodmanj · · Score: 5, Interesting

      No, the data you link to is the data that's been massaged. Specifically, it's been highly averaged. The actual AVHRR data has much better resolution. Play around with the plotting tools at ESRL : http://www.esrl.noaa.gov/psd/cgi-bin/db_search/DBSearch.pl?Dataset=NOAA+Interpolated+OLR&Variable=Outgoing+Longwave+Radiation

      That's exactly what I *did* use to create my figure. Though I had to use uninterpolated OLR data to get March 2011 data. Both data sets we've linked to are at 2.5 degree resolution. That doesn't prove that the paper's authors don't have access to higher resolution data, but no high-res data is available at the link they cite, and, I find it extraordinarily suspicious that their little blobs of peak OLR are spaced at exact multiples of 2.5 degrees apart, and lie exactly on the grid boxes for the ESRL data.

      You can generate figures for yourself that match the article's figures very neatly.

      No I cannot. Or rather, I can, but only by engaging in statistical and graphical flimflammery. You try it.

      As for the rest of your points:
      1: Yes, contentious, but I'm quoting the geology party line here. The extraordinary claim is that despite seismological evidence to the contrary, earthquakes are preceded by warning signs: that claim is the one which requires extraordinary proof.
      2: Very clear. The fault in question is in 7 km of water, close to a gigapascal of pressure. Because of Henry's Law, you don't have gaseous bubbles of anything at that pressure: all gases are in liquid solution. Thus, the gas molecules move with the water. Which is sloooowly.
      3: The figures do not match the expected behavior of a plume of material released from a point source on the Japanese coast.

      Oh, while we're quoting figures in the article, how about Figure 3, which show OLR "events" in Tohoku which are as large or larger than the ones they're interested in, occuring on Feb 22, 2011, and Jan 28, 2010. These are ignored because they're not larger than the error bars. But these error bars are bullshit: do we really believe that the natural variability of weather on March 9 is one sixth as much as on Feb 24? I sure don't. They're computing standard deviations using 6 data points, which is a recipe for disaster.