Changing Earth Rotation
from Gerhard Holtkamp, 15. March 2011, 01:42
Some news reports during the last few days mentioned that the Earth axis had shifted and the rotation slowed down due to the large earthquake in Japan on March 11, 2011. What's behind this story?
"March 11, 2011. Our first estimates of the megaquake of Honshu, Northern Japan Coast, at 5:46 UT on polar motion. According to the preliminary USGS moment tensor solution and Dahlen's dislocation model (1973), the principal axis of inertia (or figure axis) was displaced by about 14 cm at the earth surface in the direction 135° East. The effect is larger than for Chili (February 2010) and Sumatra (Dec. 2004) earthquakes. This could be observed as a step in the so-called excitation function, deduced from the determination of pole coordinates by space geodesy. But such a step could be hardly discernible from common hydro-meteorological processes."
In order to understand this news announcement by Christian BIZOUARD from the IERS EOP Center (IERS = International Earth Rotation Service) you need some background information.
In order to know how the Earth rotates in space two reference systems are needed: A celestial one and a terrestial one.
The International Celestial Reference Frame (ICRF) is mainly based on precise positions of extra-galactic radio sources (quasars, which are so far away that for all practical purposes they can be considered fixed in the sky). These positions are determined using Very Long Baseline Interferometry (VLBI). Links to the solar system are provided via precision observations of planetary spacecraft and lunar laser ranging (LLR).
The International Terrestial Reference Frame (ITRF) is an Earth-fixed reference system. Its center is the center of mass of the Earth (as determined by laser-ranging observations). The orientation of the ITRF is defined by the IERS Reference Pole (IRP) and Meridian (IRM).
The International Earth Rotation Service (IERS) is in charge of maintaining the ITRF. In principle the system is realized via the positions of reference sites. But the Earth crust isn't fixed and the reference stations keep moving with regard to each other. This means that the maintenance of the ITRF includes monitoring of the tectonic motion of the reference sites.
Thus to precisely determine the orientation and rotation of the Earth with regard to the celestial reference frame coordinated observations on a global scale with VLBI, LLR and satellite techniques (GPS, laser ranging of geodetic satellites etc.) are needed.
The Earth is not a perfect sphere but is flattened at the poles and larger at the equator. The gravitational tug of the Sun and the Moon (and to a much smaller effect the planets) on the "equatorial belly" of the Earth causes a precession of the Earth axis with a period of about 26000 years (so that our present pole star Polaris will move away from its North position over the next few thousand years) and a short-period nutation. These effects can be predicted with high precision.
Another astronomical effect on the rotation of the Earth is caused by tidal friction from the Moon and Sun. This causes the rotation to slow down so that our days become longer over time.
The axis of rotation of the Earth is not fixed compared to the Earth's crust. It seems to perform a quasi-circular path around the poles with a radius of no more than 9 meters. This so called Polar Motion can be modeled as a superposition of a 1 year period due to seasonal transport of atmospheric masses and a 432 day period (Chandler oscillation).
The Polar Motion is also affected by unpredictable geophysical forces due to the movement of masses within the Earth's interior. It is here that we finally return to earthquakes.
During Earthquakes a redistribution of mass within the Earth takes place. This modifies the "tensor of inertia" of the Earth and can affect the axis of rotation as well as the speed of rotation of the Earth (the latter due to the conservation of angular momentum).
Theoretical calculations have shown that a major earthquake can cause a shift in Polar Motion by a few centimeters and a change in the duration of the day by a few microseconds. However the changes in Polar Motion due to the transport of atmospheric and oceanic masses (in other words the weather and ocean currents) are about one or two orders of magnitude higher and would normally mask the earthquake effect unless some kind of amplification takes place.
On December 26, 2004 a devastating earthquake and Tsunami happend off the coast of Sumatra. The measurements of the associated seismic waves allowed the calculation of a so called "moment tensor" which in turn allowed the modelling of the interior Earth movements and the associated changes to the Polar Motion. The effect amounted to about 3 cm.
The diagram showing the Polar Motion as determined by the IERS between January 2003 and February 2005 does show a tiny jump but it was concluded that this was not clearly separable from the atmospheric signal and it was concluded that the effect is not discernible.
A similar calculation concerning the large earthquake in Chile on February 27, 2010, resulted in a Polar Motion of 8 cm but the mean error in modelling was given as 50 cm. A calculated decrease in the length of the day by 2 microseconds was contrasted with a mean error of 10 microseconds and daily variations due to atmospheric movements of 50 microseconds. The conclusion: The effect was not detectable.
Once again a "moment tensor" has been calculated from the measured seismic waves. Respective modelling then showed that the place where the Earth axis meets the surface at the poles has shifted by 14 cm and most likely this is what has actually happened. However once again the effect seems to be too weak to be detected unequivocally.
So the news reports about a shift in Earth's axis after the Japanese megaquake seem to be correct. But keep in mind that they are based on theoretical modelling.
"March 11, 2011. Our first estimates of the megaquake of Honshu, Northern Japan Coast, at 5:46 UT on polar motion. According to the preliminary USGS moment tensor solution and Dahlen's dislocation model (1973), the principal axis of inertia (or figure axis) was displaced by about 14 cm at the earth surface in the direction 135° East. The effect is larger than for Chili (February 2010) and Sumatra (Dec. 2004) earthquakes. This could be observed as a step in the so-called excitation function, deduced from the determination of pole coordinates by space geodesy. But such a step could be hardly discernible from common hydro-meteorological processes."
In order to understand this news announcement by Christian BIZOUARD from the IERS EOP Center (IERS = International Earth Rotation Service) you need some background information.
Orientation and Rotation of the Earth
In order to know how the Earth rotates in space two reference systems are needed: A celestial one and a terrestial one.
The International Celestial Reference Frame (ICRF) is mainly based on precise positions of extra-galactic radio sources (quasars, which are so far away that for all practical purposes they can be considered fixed in the sky). These positions are determined using Very Long Baseline Interferometry (VLBI). Links to the solar system are provided via precision observations of planetary spacecraft and lunar laser ranging (LLR).
The International Terrestial Reference Frame (ITRF) is an Earth-fixed reference system. Its center is the center of mass of the Earth (as determined by laser-ranging observations). The orientation of the ITRF is defined by the IERS Reference Pole (IRP) and Meridian (IRM).
The International Earth Rotation Service (IERS) is in charge of maintaining the ITRF. In principle the system is realized via the positions of reference sites. But the Earth crust isn't fixed and the reference stations keep moving with regard to each other. This means that the maintenance of the ITRF includes monitoring of the tectonic motion of the reference sites.
Thus to precisely determine the orientation and rotation of the Earth with regard to the celestial reference frame coordinated observations on a global scale with VLBI, LLR and satellite techniques (GPS, laser ranging of geodetic satellites etc.) are needed.
Astonomical Effects on Earth Rotation
The Earth is not a perfect sphere but is flattened at the poles and larger at the equator. The gravitational tug of the Sun and the Moon (and to a much smaller effect the planets) on the "equatorial belly" of the Earth causes a precession of the Earth axis with a period of about 26000 years (so that our present pole star Polaris will move away from its North position over the next few thousand years) and a short-period nutation. These effects can be predicted with high precision.
Another astronomical effect on the rotation of the Earth is caused by tidal friction from the Moon and Sun. This causes the rotation to slow down so that our days become longer over time.
Terrestial Effects on Earth Rotation
The axis of rotation of the Earth is not fixed compared to the Earth's crust. It seems to perform a quasi-circular path around the poles with a radius of no more than 9 meters. This so called Polar Motion can be modeled as a superposition of a 1 year period due to seasonal transport of atmospheric masses and a 432 day period (Chandler oscillation).
The Polar Motion is also affected by unpredictable geophysical forces due to the movement of masses within the Earth's interior. It is here that we finally return to earthquakes.
Earthquakes and Earth Rotation
During Earthquakes a redistribution of mass within the Earth takes place. This modifies the "tensor of inertia" of the Earth and can affect the axis of rotation as well as the speed of rotation of the Earth (the latter due to the conservation of angular momentum).
Theoretical calculations have shown that a major earthquake can cause a shift in Polar Motion by a few centimeters and a change in the duration of the day by a few microseconds. However the changes in Polar Motion due to the transport of atmospheric and oceanic masses (in other words the weather and ocean currents) are about one or two orders of magnitude higher and would normally mask the earthquake effect unless some kind of amplification takes place.
The Sumatra and Chile Earthquakes
On December 26, 2004 a devastating earthquake and Tsunami happend off the coast of Sumatra. The measurements of the associated seismic waves allowed the calculation of a so called "moment tensor" which in turn allowed the modelling of the interior Earth movements and the associated changes to the Polar Motion. The effect amounted to about 3 cm.
The diagram showing the Polar Motion as determined by the IERS between January 2003 and February 2005 does show a tiny jump but it was concluded that this was not clearly separable from the atmospheric signal and it was concluded that the effect is not discernible.
A similar calculation concerning the large earthquake in Chile on February 27, 2010, resulted in a Polar Motion of 8 cm but the mean error in modelling was given as 50 cm. A calculated decrease in the length of the day by 2 microseconds was contrasted with a mean error of 10 microseconds and daily variations due to atmospheric movements of 50 microseconds. The conclusion: The effect was not detectable.
The Earthquake on March 11, 2011
Once again a "moment tensor" has been calculated from the measured seismic waves. Respective modelling then showed that the place where the Earth axis meets the surface at the poles has shifted by 14 cm and most likely this is what has actually happened. However once again the effect seems to be too weak to be detected unequivocally.
So the news reports about a shift in Earth's axis after the Japanese megaquake seem to be correct. But keep in mind that they are based on theoretical modelling.
