google possibly predicts earthquakes

[linux18]

I've been messing around with google trends lately and noticed that the search volume for 'earthquake' increases 24-48 hours before the actual event in the region that will be effected. here are three examples:

Location: Baja California, Mexico
Earthquake happend on April 4th, search volume almost doubled on April 3rd:

http://www.google.com/trends?q=earthquake&ctab=0&geo=mx&geor=mex.bcn&date=2010-4&sort=0


Location: US (hati had too few search results)
Earthquake happened on January 12th, search volumed doubled on January 10th:

http://www.google.com/trends?q=earthquake&ctab=0&geo=us&geor=all&date=2010-1&sort=1


Location: Chile
A major 8.8 magnitude earthquake strikes on Feburuary 27, search volume up tenfold on February 26:

http://www.google.com/trends?q=earthquake&ctab=0&geo=cl&geor=all&date=2010-2&sort=0


very interesting

[lombaardcj]

I've got a theory based on what you said:

Just like animals can "sense" that such an event is likely to happen, I think people also might have a sense or "feeling" that something is about to happen. Their must be research already conducted in this field.

[Brunellus]

Correlation is not causation.

[koleoptero]

I've been messing around with google trends lately and noticed that the search volume for 'earthquake' increases 24-48 hours before the actual event in the region that will be effected. here are three examples:

Location: Baja California, Mexico
Earthquake happend on April 4th, search volume almost doubled on April 3rd:

http://www.google.com/trends?q=earthquake&ctab=0&geo=mx&geor=mex.bcn&date=2010-4&sort=0


Location: US (hati had too few search results)
Earthquake happened on January 12th, search volumed doubled on January 10th:

http://www.google.com/trends?q=earthquake&ctab=0&geo=us&geor=all&date=2010-1&sort=1


Location: Chile
A major 8.8 magnitude earthquake strikes on Feburuary 27, search volume up tenfold on February 26:

http://www.google.com/trends?q=earthquake&ctab=0&geo=cl&geor=all&date=2010-2&sort=0


very interesting
Your first and last examples nullify your statement as you can see that tiil the previous two days searches were about 0. Don't look at the lines, it's just linking the values for every day's hits there.

Correlation is not causation.
And this.