Return Pre Home Next



DURING 19992002

XU Wenyao

Institute of Geology and Geophysics, Chinese Academy of Sciences, Beijing, China

MA Shizhuang

Graduate School, Chinese Academy of Sciences, Beijing, China


Developments of geomagnetic science in China during the period of 1999-2002 are briefly summarized in following aspects: magnetic charts and models of China for 2000, main geomagnetic field and its secular variation, spatial and temporal characteristics of transient magnetic field, and possible correlation between magnetic variations and earthquakes.

Key words:  Geomagnetism, Main magnetic field, Transient magnetic field, Seismomagnetic effect


The magnetic field is one of important physical properties of the Earth. The main magnetic field originates from hydromagnetic processes in the liquid outer core, and its configuration and secular variations also depend upon the structure and dynamic processes of the core-mantle boundary, and even probably are affected by structure and processes in the mantle and lithosphere. The transient variations in the Earth's magnetic field, caused by currents in the solar-terrestrial system (mainly in the ionosphere and magnetosphere), are correlated with space weather, especially, disastrous events. Consequently, the Earth's magnetic field abounds with information of the Earth's interior and space environment.

Geomagnetism, as a part of geophysics, has steadily developed in China during the four-year period, because its importance in both understanding of the human being environment and its applications in various fields, such as navigation, resources exploration, radio communication, and space exploration.

In this paper a brief summary of researches on geomagnetism carried out by Chinese scientists during 1999-2002 is given in following respects: magnetic charts and models of China for 2000, main geomagnetic field and its secular variation, spatial and temporal characteristics of transient magnetic field. Besides, the studies on possible relationship between magnetic variations and earthquake occurrence are also summarized.


Field survey of the magnetic field was carried out during 1998-2000. Three components were measured at 158 stations, including 119 repeat stations and 39 permanent magnetic observatories, which are fairly well distributed all over the country. The magnetic charts of China for 2000 were compiled by using these data after reducing to a common time and removing solar daily variation, disturbances, and secular variations. In order to improve the boundary condition, the International Geomagnetic Reference Field for 2000 (IGRF 2000) was used to create the magnetic values at 20 positions near the boundary regions, where there are no stations.

The mathematic models of the field were established by using Taylor polynomial. The model with truncation level 4 was selected to represent the magnetic field of China for 2000 after analyzing the errors for different truncation levels. The root-mean-square errors of the models are 11.88 for declination, 13.56 for inclination, and 131.6 nT for total intensity, respectively.

The secular variation models for 2000-2005 were also established, which are represented by a Taylor polynomial with truncation level 4.

Based on geomagnetic survey and the Taylor polynomial models of China, An (2001) discussed effects of truncation level on the models, and proposed two criteria of determining proper truncation level. He also discussed boundary effects of regional magnetic field model, and suggested a technique to minimize the boundary effects.

Besides Taylor polynomial models, Rectangular Harmonic model and Cape Harmonic model for East Asia were also established (Wang et al., 1999; An and Rotanova, 2002).

Wang et al. (1999) studied the magnetic anomaly in East Asia, and established a regional residual model using the method of Rectangular Harmonic Analysis. Three magnetic components measured at 240 repeat stations in China and neighbor regions were used to set up the model with the truncation level 6. The root-mean-square errors are 134.3, 120.9, and 135.7 nT for X, Y, and Z components, respectively. They indicated that several local anomalies seem to be related with geological structures.

An and Rotanova (2002) analyzed characteristics of the magnetic anomaly in East Asia, and constructed a Cape Harmonic model with truncation level 10 on the basis of the residual magnetic field (the differences of the measured values at repeat stations and IGRF values). The root-mean-square errors are 131.2, 112.6 and 138.7 nT for X, Y, and Z components, respectively. These errors are lightly less than those in the Taylor polynomial model with truncation level 7 (133.0, 107.4, and 148.0 nT).

The magnetic anomalies in China were examined by comparing the measured components at 30 permanent observatories and model values calculated from IGRF (Wang, 2002). The results show that large deviations occur near Beijing and Shanghai. The average error for other regions is 146.9 nT.


Based on the seventh generation of International Geomagnetic Reference Field (IGRF) for 1900-2000, An and Wang (1999) and Xu, Wei and Ma (2000) examined the secular variations of the dipole moment, non-dipole components and the planetary-scale magnetic anomalies in East Asia, Oceania, South Atlantic, Africa and North America. The results show that the Earth's main magnetic field has changed dramatically during the 20th century: its dipole moment has decreased for 6.5% since 1900, the strengths of its quadrupole and octupole have increased for 95% and 74%, respectively, and the major planetary-scale magnetic anomalies on the Earth's surface have enhanced for 21% -56%; besides, the magnetic center has shifted for 200 km towards Pacific Ocean. These time-variation features appear to be similar to behaviors before a geomagnetic polarity reversal.

The characteristics of a planetary-scale geomagnetic anomaly may be described by ‘unsigned magnetic flux' through the whole anomaly region. The eighth generation of IGRF was analysed to study the secular variations of the planetary-scale geomagnetic anomalies during the period of 1900-2000 (Xu, 2001a). The magnetic fluxes through the Southern Atlantic, Australian, and African anomalies have increased by more than 200 MWb; the increase of the flux through the Eurasian anomaly is smaller (157 MWb), while the flux through the North America anomaly has decreased by 50 MWb.

Study on geomagnetic energy distribution in the Earth's interior and its secular variations shows that during the interval of 1900-2005 the total geomagnetic energy beyond the core-mantle boundary (CMB) decreases by 3.3%, while the magnetic energy above the Earth's surface decreases in this period by 11.4% (Xu, 2001b). The magnetic energy in different layers varies with time in different way. The amounts of the energy in the crust, upper mantle, transition belt and lower mantle decrease, however, the energy in the D” layer increases rather rapidly.  The time variations of the energy density at different depths indicate a clearly demarcation at r = 3840 km, above and below which the energy decreases and increases, respectively.  The characteristics mentioned above suggest that as the total energy of the geomagnetic field steadily decreases, a redistribution of the energy takes place in the Earth's interior, and the magnetic energy is gradually concentrates towards the bottom of the mantle, mainly the layer D”. The rapid increase of the magnetic energy in the D” may be ascribed to growth of high-order multipoles in the geomagnetic field, which resembles, to a certain extent, the behaviors prior to a geomagnetic polarity transition when the dipole moment decreases and at same time the multipoles intensify.

Under the assumption of insulated mantle, the IGRF models are used to calculate the geomagnetic field structure in the Earth' interior, from the ground surface to the Core-Mantle Boundary (CMB) (Xu and Wei, 2001). Four reversed-polarity patches are revealed, they are the Southern Africa (+Z) and Southern America (+Z) in the Southern Hemisphere, North Polar region (-Z) and Northern Pacific region (-Z) in the Northern Hemisphere. Extending upward, each of the reversed polarity patches at the CMB forms a chimney-shaped “reversed polarity column” in the mantle with the bottom at the CMB. The height of the SAF column has grown rapidly from 200 km in 1900 to 900 km in 2000. If the column grows steadily at the same rate in the future, its top will reach to the ground surface in 600700 years. And then a reversed polarity patch will be observed at the Earth's surface, which will be an indicator of the beginning of a magnetic field reversal. 

The IGRF model series for 19002000 is too short to study the geomagnetic secular variation, especially for long-period variations. A longer series of the geomagnetic field models is needed. On the basis of the principle of natural orthogonal components analysis (NOC), Xu 2002a, b, c constructed “NOC model of the geomagnetic field”, and developed a new technique, by which a magnetic model with low truncation level can be upgraded to a higher truncation level.

Using the main-field models IGRF 1900-2000, Xu 2002a calculated the principal components of the magnetic field. Each of the components has specific spatial structure and varying intensity.  The structures of the principal components are rather stable for different periods. 

This method is first used to revise the high-degree Gauss coefficients (n >= 7) in the IGRF models for 19451955 Xu, 2002b, which exhibits some unusual and unreliable behaviors in comparison with the models for other epochs (Xu, 2000). These irregular variations have little effect on the main features of the surface magnetic pattern. However, when we extrapolate the field pattern downward through the insulating mantle to the core-mantle boundary (CMB), the contributions of the higher-degree coefficients become more important and are likely to affect the shape of the geomagnetic energy spectrum, distribution of magnetic flux, and magnetic determination of the conducting core radius. The revised high-degree coefficients exhibit fairly smooth time-variations. 

Taking the model for 1885 (truncation level ) as an example and solving a linear equation system, Xu 2002c calculated the high-degree Gauss coefficients (n = 710) from the existing low-degree coefficients (n = 16). Adding this model to the model series of IGRF 19002000, a longer model series for 18852000 with a common truncation level is obtained.

The westward drift is one of the most important features in the main magnetic field. Using method of Correlation Analysis of Moving Random Pattern (CAMRAP), Wei and Xu 2000, 2001a,b studied westward drift of the non-dipole fields in IGRF 19001995. The results show that the anomaly of Asia-Europe Continent has drifted with an average rate 0.07°/a. Particularly, during 19001930 the westward drift is quite steady, then the drift is slow down during 19301980, and eventually the drift turned to eastward after 1980. Similar study is made for other magnetic anomalies. The anomaly in South Atlantic Ocean has steadily drifted westward at an average rate 0.13°/a  with a remarkable pattern deformation.

In contrast with the fast westward drift (0.2°/a) on the Earth surface, the westward drift at the core-mantle boundary (CMB) is much slower, less than 0.1°/a (Xu and Wei, 2001). This contrast may be attributed to the different westward drift velocities of different harmonics in the geomagnetic field.


Transient variations in the geomagnetic field, covering a broad frequency band from short period micropulsation to 11-year solar cycle variation, arise from the currents in the ionosphere and magnetosphere, as well as induced currents in the Earth's interior. Consequently, their characteristics depend on both the near-Earth space environment, such as conditions of the solar wind and the interplanetary magnetic field (IMF), and the electromagnetic properties in the Earth (Gao and Zhang, 2000, Xu, 2000b, Liu S.-L. and Li L.-W., 2002).

1. Magnetic Storm and Substorm

Magnetic storm and substorm are most important types of transient magnetic variations, which are closely related with solar-terrestrial energy coupling, especially, coronal mass ejection (CME) and interplanetary shock. They concern the variations in the electromagnetic field, electric currents, plasma convection, and plasma waves in the ionosphere and magnetosphere (Gao and Zhang, 2000, Wang and Feng, 2002, Liu S.-L. and Li L.-W., 2002, Liu et al. 2002). 

The magnetic records of So-Se Observatory near Shanghai, China from the year 1877 down to the present, of more than hundred years long, afford a very good utilization of making investigation on magnetic storms and substorms. Using this long series of dataset, Tschu (2001) extensively discussed the scaling of K index during 19221957, which is for the first time a long series of local K index from a low latitude observatory in the Far East. He statistically analyzed 2498 geomagnetic storms during 19081976 and 1282 bay disturbances during 19331953. Physical interpretation was discussed using the present understanding of geomagnetic disturbances together with the solar-terrestrial relationship.

The latitudinal effect of magnetic storm was studied using the data from a low-latitude meridian chain of magnetometers in China, named SMALL established by Chinese and American scientists (Gao et al., 2000, Yang et al., 2002). The results show that both the range of H component for sudden commencement and the initial phase duration decreases with decreasing latitude, while the depression of H component during the main phase increases with decreasing latitude.

High-latitude meridian chains of magnetometers supply a powerful tool to study the coupling process between the magnetosphere and ionosphere. Using the magnetic records with 1-minute time resolution at 22 stations of two meridian chains, Shen et al. (1999a, b) examined the latitudinal dependence of magnetic variations during magnetic storms and associated ionospheric disturbances. It is shown that the geomagnetic response of the magnetosphere-ionosphere coupling is local time dependent, which is caused by combined effects of the magnetospheric convection pattern and field-aligned currents.

2.  Magnetic Disturbances and Currents in the Ionosphere-Magnetosphere System

Transient magnetic fields recorded at the magnetic stations and spscecraft are superposition of the magnetic fields caused by various current systems in the ionosphere and magnetosphere, which are associated with different physical processes. It seems to be necessary to separate magnetic contributions from each of the current systems for studying the magnetic field origin and associated physical processes.

The method of natural orthogonal components (NOC) was used to separate the driven and

loading-unloading processes during substorms (Xu and Sun, 2000).

The magnetic variation at mid-low latitudes is mainly caused by ionospheric dynamo currents, ring currents and field-aligned currents. Chen and Xu (2001) constructed the theoretical models for each of these currents, and calculated their magnetic fields at the ground surface. Comparing the measured magnetic fields during quite condition and disturbed periods with the calculated model values, they separated the magnetic fields caused by each of the current systems. Using this technique, Chen et al. (2001) studied the evolution of the great magnetic storm occurred on July 1516, 2000.

Comparing the current systems in geomagnetic coordinate system and corrected geomagnetic coordinate system indicates that the conjugate features of north and south polar regions are shown much more clearly in the corrected geomagnetic coordinate system (Chen et al., 2000a, b).

The magnetic records from high-latitude meridian chains of magnetometers are used to study evolution of the current systems during the magnetic storm of July 1516, 2000 (Chen et al., 2001).

3.  Magnetic Micropulsation (ULF)

Magnetic micropulsation, as an important type of transient variations in the geomagnetic field, has been studied by using the data obtained at mid-low latitudes and in Antarctica. Yang et al. (1999) compared the magnetic pulsations recorded at Zhongshan Station of East Antarctica and Changcheng Station of West Antarctica during a great magnetic storm, and indicated the differences between dayside- and nightside-pulsations in initial, main and recovery phases. Various types of the magnetic pulsations during storms were examined by Wang and Feng (2001).

The characteristics (such as period, power spectrum, polarization etc.) of Pi2 micropulsation were studied by using the data from Zhongshan station, Antarctica (Du et al, 1999). The results show that Pi2 micropulsation usually occur in night side (20:00 – 02:00 MLT), its main frequency band is between 6.79 mHz and 1.35 mHz, and this band becomes narrower around midnight; the polarization of Pi2 is almost linear. An integrated analysis on Pc5 and Pc3 in the cusp region was made by Liu et al. (1999, 2001a, b). Using cross-spectrum technique to the ULF data simultaneously recorded at Zhongshan and Davis stations in Antarctica, they obtained the principal propagation direction of Pc5: in daytime Pc5 propagates westward in the morning and eastward in the afternoon; while in nighttime Pc5 propagates westward before 20:00 MLT, and eastward after this local time (Liu et al., 2001a).  Same analysis was made for Pc3. It is shown that a westward propagation is dominant in daytime (Liu et al., 1999, 2001b, Yang and Liu, 1999, Yang, 2000).

The east-west propagation characteristics of Pi2 and Pc3 pulsations at low latitudes were studied by means of a latitudinal chain of magnetometers in northern China (Yang, et al., 1999, 2000). The general features are similar to Antarctica. Namely, the westward propagation is dominant in the morning and the eastward propagation is dominant in the afternoon

Analyzing the polarization of Pc3-4 recorded at low latitudes during 3 eclipses, Wang et al. (1999) indicated that the polarization changes by 90°if the station is located equatorward to the total eclipse zone, while there are no any changes in polarization if the station is poleward to the total eclipse zone.  This feature implies a filtering mechanism.


A lot of events suggest that geomagnetic field variations are perhaps associated with earthquakes or other natural phenomena. This probable correlation leads some researchers to find out their connection and the physical mechanism. 

Seismomagnetic effects have been extensively studied since the great earthquakes at Xingtai in 1966, at Haicheng in 1973 and at Tangshan in 1976. The interval between successive magnetic storms, transfer function of horizontal and vertical components for different frequencies, anomalous feature of daily variation, and other precursors in magnetic field have been used to predict coming earthquakes.

In recent years some of Chinese scientists focus on abnormal behaviors of ULF electromagnetic emission, or magnetic pulsation, before great earthquakes.  Several events show that a few days before earthquakes, ULF electromagnetic emission is enhanced, and its polarization varies systematically (Yang and Du, 2001). Analyzing the ULF data recorded at a latitudinal chain of magnetic stations before the Hetian (northwest China) earthquake (ML=7.1) occurred on Nov. 19, 1996, Du et al. (2002) indicated a close relationship between the orientation of the epicenter and polarization direction of ULF electromagnetic emission in the frequency band 0.08350.167 Hz before the earthquake.    


An Z.-C. and Wang Y.-H. (1999), Global changes of the non-dipole magnetic fields for 1900-2000, Chinese J. Geophys., 42(2):169-177.

An Z.-C. (2001), Analysis and discussion of the geomagnetic field polynomial models, Chinese J. Geophys., 44(supplement)45-50.

An Z.-C, Rotanova,N.M. (2002),  Calculations and analysises of the geomagnetic field models for East Asia, Chinese J. Geophys., 45(1)34-41.

Chen G.-X., Du A.-M. , Xu W.-Y, Chen H.-F., Hong M.-H, Peng F.-L, and Shi E.-Q. ( 2001), Response of high latitude magnetic field to the magnetic storm of July 15-16, 2000, Science in China (A), 31( supp), 112-119.

Chen H.-F., Chen G.-X, Peng F.-L, and Xu W.-Y (2000a), Analysis of  current system by using corrected geomagnetic coordinates, Chinese J. Polar Sci., 11(1), 59-66.

Chen H.-F, Chen G.-X., Peng F.-L., and Xu W.-Y. (2000b), Polar electrojet during quite period in corrected geomagnetic coordinates, Sci. in China (A), 30(supp.), 88-91.

Chen H.-F. and Xu W.-Y. (2001), Variations of inner magnetospheric currents during the magnetic storm of May 1998, Chinese J. Geophys., 44(4)490-499.

Chen H.-F., Xu W.-Y., Chen G.-X., Hong M.-H., and Peng F.-L. (2001), Latitudinal characteristics of the geomagnetic field during the storm of July 15-16, 2000, Solar Phys., 204, 339-349.

Du A.-M., Yang S.-F., and Xiao F.-H. (1999), The characteristics of Pi2 pulsations at Zhongshan station of Antarctica, Chinese J. Geophys., 10(2), 171-175.

Du A.-M., Huang Q.-H., and Yang S.-F. (2002), Epicenter location by abnormal ULF electromagnetic emission, Geophys. Res. Lett., 29(10), 1029-1031.

Gao Y.-F., Chi P.-J.,  Le, G. Russell C.T., Yang D.-M., Zhou, X., Yang S.-F., Angelopoulos V., and Chun F.-K. (2000), Sino-magnetic array at low latitude (SMALL), including initial results from the sister site in United States, Adv. Space Res., 25(7/8), 1343-1351.

Gao Y.-F. and Zhang X.-L. (2000), Relations between IMF Bs events and intense magnetic storms,  Chinese J. Space Sci., 20(2), 136-143.

Liu S.-L. and Li L.-W. (2002), Study on relationship between southward IMF events and geomagnetic storms, Chinese J. Geophys., 45(3), 297-305.

Liu S.-L., Guo J.-G., Zong Q.-G., Wilken B., Fu S.-Y. (2002), The magnetic storm resulted from the interplanetary shock of February 21, 1994, Chinese J. Space Sci., 22(3), 203-211.

Liu Y.-H., Liu R.-Y., Yang S.-F, He L.-S, and Fraser B.J. (1999), Propagation characteristics of Pc3 frequency rang pulsations in the cusp latitude, Chinese J. Polar Sci., 10(2),163-170

Liu Y.-H., Liu R.-Y., Yang S.-F., He L.-S., and Fraser B. J. (2001a), Propagation and origin of Pc5 frequency range pulsation in the cusp latitude, Chinese J. Geophys., 44(supplement)8-15.

Liu Y.-H., Liu R.-Y., Yang S.-F., He L.-S., and Fraser, B. J. (2001b),  A study of Pc3 pulsation in the cusp latitudes by short based-line, Chinese J. Geophys., 44(supplement)9-21.

Shen C.-S., Zi M.-Y., Gao Y.-F., Suo Y.-C., and Wu J. (1999a), Geomagnetic responses to the convection field, field-aligned current and auroral electrojet, Chinese J. Geophys., 42(6)725-731.

Shen C.-S., Zi M.-Y., Gao Y.-F., Suo Y.-C., and Wu J. (1999b), On features of the manetospheric coupling examined by geomagnetic data with high time resolution, Chinese J. Space Sci., 19(2)134-140.

Tschu, K. K. (2001), Studies on the disturbance variations of geomagnetic field at So-Se, near Shanghai, China, Chinese J. Geophys., 44(supplement)51-67.

Wang, D.-J., Chen S.-W., and Zhou G.-C. (1999), Source of the Pc3-4 geomagnetic pulsation in the very low latitude region studied by eclipse effects of geomagnetic pulsation, Chinese J. Geophys., 42(4), 460-464.

Wang T.-W. and Feng X.-S. (2002), The geomagnetic response of May 1998 event, Chinese J. Space Sci., 21(2), 157-164.

Wang T.-W. (2002), The analysis of the IGRF error in the China Continent, Chinese Journal of Geophys., ( in press).

Wang Y.-H, An Z.-C, Golovkov, V. P., Rotanova, N. M., and Kharitonov, A. L. (1999), Theoretical analysis of geomagnetic field over East Asia and rectangular harmonic model, Chinese J. Geophys., 42(5), 640-647.

Wei Z.-G. and Xu W.-Y. (2000),  Westward drift of the geomagnetic anomaly in East Asia, Chinese J. Geophys., 43(1), 49-56.

Wei Z.-G. and Xu W.-Y. (2001a),  Drift and intensity variations of the geomagnetic field, Chinese J. Geophys., 44(4), 500-109.

Wei Z.-G and Xu W.-Y. (2001b),  Latitudinal dependence of westward drift of the geomagnetic field and its dispersion, Chinese Sci. Bull., 46(18), 1563-1567.

Xu W.-Y. (2000a), Unusual behavior of the IGRF during the 1945-1955 period,  Earth Plan. Space, 52(12), 1227-1233.

Xu W.-Y. (2000b), Effect of geomagnetic field configuration on space weather, Sci. in China (A), 21-24.

Xu W.-Y. and Sun W., A study on the multi-component substorm current, Chinese J. Polar Sci., 11(1), 53-58, 2000.

Xu W.-Y, Wei Z.-G, and Ma S.-Z. (2000), Dramatic variations in the Earth's main magnetic field during the 20th century, Chinese Sci. Bull., 45(21), 2013-2016.

Xu W.-Y. (2001a),  Secular variations of the planetary-scale geomagnetic anomalies, Chinese J. Geophys., 44(2), 180-189.

Xu W.-Y. (2001b), Distribution of geomagnetic energy in the Earth's interior and its secular variation, Chinese J. Geophys., 44(6)747-753.

Xu W.-Y and Wei Z.-G. (2001), Reversed polarity patches at core-mantle boundary and geomagnetic field reversal, Sci. in China (D), 31(8)617-625.

Xu W.-Y. ( 2002a), NOC model of the geomagnetic field, Science in China (D), 31(7), 576-587.

Xu W.-Y. (2002b), Revision of the high-degree Gauss coefficients in the IGRF 1945-1955 models by using natural orthogonal component analysis, Earth Plan. Space, 54 ( in press).

Xu W.-Y., 2002c,  Natural orthogonal component analysis of IGRF and its application to study on the historical geomagnetic models, Geophys. J. Int., 143 (in press)

Yang S.-F, Liu Y.-H. ( 1999), Polarization characteristics of Pc3 pulsations at Zhongshan station of Antarctica,  Chinese J. Polar Sci., 10(2), 155-162.

Yang S.-F., Xiao, F.-H., and Du A.-M. (1999), Analysis of geomagnetic pulsations in Antarctic region during magnetic storm on March 24, 1991, Chinese J. Geophys., 42(3)310-321.

Yang S.-F, Du A.-M, Chen B.-S,, Ning X.-R, Xiao F.-H. Hong F.-Y. (1999), Propagation characteristics of low-latitude Pi2 pulsations in east-west direction,  Chinese J. Space Sci., 19(3), 232-238.

Yang S.-F. (2000),  Digital filter technology and its application to geomagnetic pulsations in Antarctica,  Chinese J. Polar Sci. , 11(1), 67-73.

Yang S.-F, Du A.-M, Ning X.-R, Chen B.-S, Xiao F.-H. Hong F.-Y. (2000), Propagation characteristics of low-latitude Pc3 pulsations in east-west direction,  Chinese J. Geophys., 43(2), 213-222.

Yang S.-F, and Du A.-M. (2001),  Relationship between abnormal ULF electromagnetic emission and the direction of sources before earthquakes in Kashi region of Xinjiang in November, 1996, Chinese J. Geophys., 45(1)101-108.

Yang S.-F, Du A.-M, Gao Y.-F., and Han D.-S. (2002),  Latitudinal effect of magnetic storm on April 6, 2000 at low-latitude meridian chain, Chinese J. Geophys., 45(4)461-469.

Return Pre Home Next