Ionospheric Anomalies Appear Before the 12 May 2008 Mw7.9 Sichuan Earthquake J. Y. Liu1,2, Y. I. Chen3, C. H. Chen1, C. Y. Chen1, M. Nishihashi4, K. I. Oyama1, and K. Hattori4 1Institute of Space Science, 2Center for Space and Remote Sensing Research, National Central University, Taiwan 3Institute of Statistics, National Central University, Taiwan 4Graduate School of Science and Technology, Chiba University, Japan The total electron content (TEC) of Global ionospheric map (GIM) and electron density profiles probed by six micro satellites of FORMOSAT3/COSMIC (F3/C) are for the first to simultaneously observe seismo-ionospheric anomalies during an M7.9 earthquake in Sichuan, China on 12 May 2008. It is found that ionospheric electron densities above the fore coming epicenter anomalously decrease in the afternoon period of day 6 to 4 and the late evening period of day 3 before the earthquake. Results further show that the ionospheric F2 layer significantly descends about 50-80km and the GPS TEC anomalously reduces about 1650km and 3300km in radius from the epicenter in the latitudinal and longitudinal direction, respectively. An earthquake with magnitude Mw7.9 occurred in eastern Sichuan, China (30.986°N, 103.364°E, Depth 19km) at 06:28:01 UT on Monday 12 May 20081. The global ionosphere map2 (GIM) of the total electron content (TEC) constructed with 1000s of worldwide ground-based receivers of the global positioning system (GPS) 1 and routinely published in a 2-hour time interval is used to observe anomalous signatures induced by the earthquake. Here, the TEC is integrated the ionospheric electron density between a ground-based receiver and a GPS satellite at about 20200km altitude. The spatial resolutions of the GIM on the ±87.5oN latitude and ± 180oE longitude are 2.5° and 5°, respectively. Therefore, each map consists of 5040 (=70x72) grid points (Fig. 1). Similar to a Geostationary Meteorological Satellite hourly observing clouds for the metrological weather, the GIM now is employed to observe signatures of the lithospheric, atmospheric, and ionospheric weather (such as thunderstorm, ionospheric storm, and earthquake). For each grid point, we compute the median of the 1-30 days before the earthquake (12 April to 11 May 2008) and difference of each observation day to the median. Figure 1 displays the GIMs at 08:00 UT and at global fixed local time of 15:00LT on day 6 before the earthquake (6 May 2008), the associated median, and the extreme reduce difference on 6 May 2008. It is interesting to see that GPS TECs above the fore coming Sichuan epicenter on 6 May 2008 significantly and extremely decrease. Note that the median represents the background while the extreme reduction stands for possible earthquake related anomaly. We focus on temporal variations of the GPS TEC above the epicenter isolated from the GIM database during the earthquake. Figure 2 displays that the ionospheric GPS TEC significantly decrease in the afternoon period on 6, 7, and 8 May (6-4 days before the earthquake) as well as in late evening of 9 May (3 days before). To find spatial distributions of these anomalies, sequences of GIMs at 06:00 and 14:00 UT (13:00 and 21:00 LT at Sichuan) of the 30-day period (12 April to 11 May 2008) are examined. Figure 3 shows that the decrease GPS TECs and the extreme reduction 2 differences generally appear around the Sichuan epicenter at 06:00UT on 6, 7, and 8 May 2008, as well as at 14:00UT on 9 May 2008. The local time of 06:00 and 14:00UT at the Sichuan area are the afternoon period of 13:00LT and the late evening period of 21:00 LT, respectively. Taking into account local time effects, sequences of GIMs at the fixed global times of 13:00 and 21:00LT on the four days are also studied by the same process. We again confirm that the decrease GPS TECs and the extreme reductions appear around the epicenter in the afternoon period of day 6-4 and late evening period of day 3 before the Sichuan earthquake (Fig. 3). Six microsatellites of the joint Taiwan-US satellite constellation mission, termed FORMOSAT-3/COSMIC, were successfully launched in to a circle low Earth orbit at 01:40 UT on 15 April 20063. Each satellite houses a GPS occultation experiment (GOX) payload (space-based GPS receiver) applying a powerful technique of the atmospheric radio occultation4 to globally derive the vertical profile of electron density in the ionosphere. Figure 4 reveals the electron density profiles observed over the epicenter during the afternoon period of 13:00-17:00LT on 6, 7, and 8 of May and the late evening period 21:00-24:00LT on 9 May as well as their associated references of the median, lower quartile, and upper quartile of the 15-day period from 21 April to 5 of May 2008, which is day 7-22 before the earthquake. The satellite results confirm that the electron density of the ionospheric F2-peak NmF2 in the afternoon period on 6, 7, and 8 of May and in the late evening period of 9 May are significantly less than their the median and lower quartile. Moreover, it can be seen that the F2-peak height descends from about 300 km down to 200s km altitude which is approximately 50-80 km lower than the associated median, on the 4 anomalous days. 3 Large earthquakes are often preceded or accompanied by signals of different nature: electric, magnetic, electromagnetic, or luminous, although seismic waves are the most obvious manifestation5,6. Recently, ionospheric NmF2 and GPS TEC anomalous variations appeared before earthquakes have received considerable discussions7-10. The NmF2 is the greatest electron density of the ionosphere while the TEC is the integration of the ionospheric electron density, and the NmF2 and TEC generally are highly correlated11. A statistical investigation demonstrates that the abnormal decrease of the ionospheric NmF2, in term of the corresponding plasma frequency foF2, in the afternoon period of 12:00-18:00LT occurs significantly within 1-5 days before 184 M5.0 earthquakes during a 6-year period of 1994-1999 in the Taiwan area12. An extension work shows that the ionospheric GPS TEC pronouncedly reduces in the afternoon period of 12:00-18:00 LT and especially evening period of 18:00–22:00 LT within 5 days prior to 20 M6.0 earthquakes in Taiwan during September 1999 -December 200213. A study of GIM data during 160 worldwide M7.0 earthquakes in January 1999 - May 2008 confirms that the GPS TECs above the epicenters significantly decrease in the afternoon period day 1-5 before the earthquakes14. The existing results suggest that the GPS TEC decrease anomalies in the afternoon and evening periods of 6-9 May 2008 are related to the Sichuan earthquake. The statistical study of 184 M5.0 earthquakes shows that the reduction anomaly of the ionospheric electron density appears more often before larger earthquakes, but less likely away from the epicenter12. On the other hand, the spatial analyses reveal that the ionospheric GPS TECs centered on the epicenters notably decrease day 3-4 before the 20 September Mw 7.6 Chi-Chi earthquake15 and day 5 prior to the 24 4 December 2004 Mw9.3 Sumatra earthquake16. It can be seen that the GPS TECs once again significantly decrease about 10-15° in latitude and 15-30° in longitude from the epicenter in the afternoon and evening periods on day 3-6 (6-9 May 2008) before the Sichuan earthquake. Since 1° is corresponding to about 110 km, the anomalous size of the Sichuan earthquake is 1650km (=110km/ ° × 15 ° ) in the latitudinal and 3300km (=110km/°×30°) in the longitudinal direction, respectively. On the other hand, in the lithosphere the earthquake preparation area can be estimated by R =100.43M, where R is the radius of the earthquake preparation zone, and M is the earthquake magnitude17. For the Mw7.9 Sichuan earthquake, we obtain R = 2495 km, which is about the similar size in the latitudinal but two times greater in longitudinal direction of the anomaly size observed in this paper. On the other hand, the F3/C GOX observes that tremendous amount of the electron density disappears from the upper part of the ionosphere on those four anomalous days. Comparing to the associated median, the electron density of NmF2 decreases more than 30-50% and the height of HmF2 descends about 50-80 km on the anomalous days. Note that it shall need enormous power and/or energy to modify the ionosphere in the area of 1650km in the latitudinal and 3300km in the longitudinal direction. Very colorful earthquake clouds appear around the epicenter several tens minutes before the Sichuan earthquake onset18,19. This indicates that the electromagnetic environment has been significantly changed and seismoelectromagnetic signals (SESs) have been activated around the fore coming epicenter area during the earthquake preparation period. Due to abundance of electrons, the ionosphere above about 100km becomes highly conductive20. Here the lithosphere and the ionosphere could act as a concentric conducting sphere and shell, and 5 therefore any electromagnetic anomalies, such as currents, charges, vertical electric field etc., appearing on Earth’s surface generated during the earthquake preparation period can easily and quickly map to and affect the ionosphere by modifying the spatial distribution of the electron density within. Although the SESs were not directly measured, the observed colorful earthquake clouds indicate either vertical electric field and/or broadband electromagnetic emissions on the Earth’s surface being important. The dip angle of the Earth’s magnetic field around Sichuan earthquake (30.986°N, 103.364°E) is about 45°, and therefore an upward plasma flow (polar wind like) due to upward penetration electric fields20 and/or a plasma expansion due to electromagnetic emissions can easily result in the anomalous decrease of the electron density in upper part of the ionosphere during the earthquake preparation period. Acknowledgements This research was partially supported by the Ministry of Education Grant 91-N-FA07-7-4, the iSTEP project, to National Central University. 6 References 1. http://earthquake.usgs.gov/eqcenter/recenteqsww/Quakes/us2008ryan.php 2. ftp://cddisa.gsfc.nasa.gov/pub/gps/products/ionex 3. Cheng, C.-Z. F., Kuo, Y.-H., Anthes, R. A., & Wu, L., Satellite constellation monitors global and space weather, Eos Trans. AGU, 87(17), 166 (2006). 4. Yunck, T., An Overview of Atmospheric Radio Occultation, J. Global Positioning System, 1, 58-60 (2002). 5. Bolt, B. A. Earthquake, W. H. Freeman and Company, New York, ed. 4 (1999). 6. Freund, F. Time-resolved study of charge generation and propagation in igneous rocks. J. Geophys. Res. 105, 11001-11019 (2000). 7. Hayakawa, M. & Molchanov, O. A. Ed., Seismo Electromagnetics: LithosphereAtmosphere-Ionosphere Coupling, TERRAPUB Tokyo (2000). 8. Pulinets, S. & Boyarchuk, K. Ionospheric Precursors of Earthquakes, Springer Berlin Heidelberg New York (2004). 9. Kamogawa, M. Preseismic lithosphere-atmosphere-ionosphere coupling. Eos 87, 417 (2006). 10. . Rishbeth, H. Ionoquakes: Earthquake precursors in the ionosphere. Eos, 87, 316 (2006). 11. Liu, J. Y., Chen, Y. I., Jhuang, H. K., & Lin, Y. H. Ionospheric foF2 and TEC anomalous days associated with M5.0 earthquakes in Taiwan during 1997-1999. Terr. Atmo. Oce. Sci., 15, 371-383 (2004). 12. Liu, J. Y., Chen, Y. I., Chuo, Y. J. & Chen, C. S. A statistical investigation of pre-earthquake ionospheric anomaly. J. Geophys. Res., 111, A05304, doi:10.1029/2005JA011333 (2006). 13. Liu, J. Y., Chuo, Y. J., Shan, S. J., Tsai, Y. B., Chen, Y. I., Pulinets, S. A. & Yu, S. B. Pre-earthquake ionospheric anomalies registered by continuous GPS TEC measurements, Ann. Geophysicae, 22, 1585-1593 (2004). 14. Liu, C. Y., A study of seismo-ionospheric GPS TEC anomalies over the epicenters of strong earthquakes. MS thesis, Beijing University of technology, May 2008. 7 15. Liu, J. Y., Chen, Y. I., Chuo, Y. J. & Tsai, H. F. Variations of ionospheric total electron content during the Chi-Chi earthquake, Geophys. Res. Lett. 28, 13811386 (2001). 16. Liu, J. Y. & Chen, Y. I. Ionospheric precursors of the 26 December 2004 M9.3 Sumatra Earthquake, Geophys. Res. Lett. under reviewing (2008). 17. Dobrovolsky, I. P., Zubkov, S. I. & Miachkin, V. I. Estimation of the size of earthquake preparation zones. Pure and Appl. Geophys., 117, 1025-1044 (1979). 18. http://www.youtube.com/watch?v=1IHoZoAVLo0 19. http://www.youtube.com/watch?v=YZOy8Vt_go8 20. Kelley, M. C. The Earth’s Ionosphere. Elsevier, New York, 487 pp. (1989). 8 Figure legends Figure 1 | The GIMs observed 6 days before the Sichuan Earthquake. The left and right columns are GIMs at 08:00 UT and global fixed local time of 15:00LT, respectively. The top panels are observed on 6 May 2008 (day 6 before the earthquake), the second top panels are the median of the period of 12 April - 11 May 2008 (1-30 days before the earthquake), and the third top panels denote the extreme minima of the 30 day appear on 6 May 2008. The color denotes the difference of the TEC observed on 6 May 2008 from the associated 30-day median. It can be seen that the ionospheric GPS TECs around the Sichuan epicenter marked by the circle dashed circle significantly reduce. The bottom panels are the magnified of the difference between 6 May 2008 and the associated median. The red grids denote the 30-day extreme minima. The GIM grid points lie between ±87.5oN and ±180oE with 2.5° and 5° grid intervals in the latitudinal and longitudinal directions, and therefore each map has 5040 (=70x72) grid points in total. Figure 2 | A time series of GPS TEC right above the Sichuan epicenter extracted from GIMs during the 2008 M7.9 Sichuan Earthquake. The red curve is the observation while the two dark curves are the low and upper bounds. The low bound (LB) and upper bound (UB) are constructed by the 1-15 pervious days moving median (M), lower quartile (LQ), and upper quartile (UQ). Here, LB=M-1.5(M-LQ) and UB=M+1.5(UQ-M). It can be seen that the GPS TEC anomalously reduces during 06:00-10:00UT (the afternoon period of 13:00-17:00LT) on 6, 7, and 8 May as well as 14:00-17:00UT (the late evening period of 21:00-24:00LT) on 9 May 2008. The blue dots denote the extreme minima of the 30 day period before the earthquake. 9 Figure 3 | The ionospheric GPS TEC anomalies appear on the GIMs 6 to 3 days prior to the Sichuan Earthquake. The left and right columns are GIMs of the universal time and global fixed local time, respectively. The top three panels are the difference between the observations and the associated medians at 06:00UT and global fixed 13:00LT on 6, 7, and 8 May 2008 (from top to down). The color denotes the difference of the TEC on the observation day from the associated 30-day median, and the red grids represent the 30-day extreme minima. The bottom panels are at 14:00UT and global fixed 21:00LT on 9 May, 2008. It can be seen that the ionospheric GPS TECs around the epicenter significantly reduce. Figure 4 | The ionospheric electron density above the epicenter observed on 6 to 3 days before the Sichuan Earthquake by FORMOSAT3/COSMIC satellites. The panels from top to down are the electron density vertical profiles observed during 13:00-16:00LT on 6, 7, and 8 as well as 21:00-24:00LT on 9 May 2008, respectively. The red curves represent the observed profiles while the solid and dashed curves in each panel are the associated median and upper/lower quartiles at the same local time period during 21 April – 5 May 2008 (7 to 22 days before the earthquake). It confirms that the electron density significantly decreases on 6 to 3 days before the earthquakes. 10 Figure 1. 11 P P P P ←M7.9 Figure 2. 12 Figure 3 13 Figure 4 14