Ionospheric Response to the Geomagnetic Storm on August 21, 2003 ...
3212
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, VOL. 48, NO. 8, AUGUST 2010
Ionospheric Response to the Geomagnetic Storm on August 21, 2003 Over China Using GNSS-Based Tomographic Technique Debao Wen, Yunbin Yuan, Jikun Ou, and Kefei Zhang
Abstract—The impacts of the August 21, 2003 geomagnetic storm on the ionosphere over China have been first investigated by using the so-called computerized ionospheric tomography (CIT) technique and the observations of the Crustal Movement Observation Network of China. Tomographic results show that the main ionospheric effects of this geomagnetic storm over China are as follows: 1) the negative storm phase effect appears in the F region and 2) the positive storm phase effect occurs above the F region. Meanwhile, some key features in the ionospheric structure have been revealed in the ionospheric images during the storm; this includes the disturbances and an elongated region of the reduced electron density at the latitude around 32◦ N. Statistical comparisons are carried out to confirm the reliability of the global-navigation-satellite-system-based CIT reconstruction results using the profile obtained from ionosonde observations. Index Terms—Electron density, geomagnetic storm, global navigation satellite systems (GNSS), ionosphere, tomography.
I. I NTRODUCTION
T
HE OCCURRENCE of geomagnetic storms greatly changes the spatial distribution of ionization. This may also affect the propagation of radio signals and deteriorate the performance of global navigation satellite systems (GNSS). In the past, traditional ground-based instruments, such as incoherent scatter radars (ISRs) and ionosondes, have been extensively used for the investigation of the effects of geomagnetic storms on the ionosphere. However, ionosondes cannot be used to measure the topside ionosphere, and sometimes they suffer from absorption problems during geomagnetic storms, while ISRs are limited by geographical locations [1]. Given these limitations and the global impacts of geomagnetic storm, it is Manuscript received July 14, 2009; revised December 15, 2009. Date of publication April 8, 2010; date of current version July 21, 2010. This work was supported in part by the National Science Foundation of China under Grants 40804002 and 40890160, by the Scientific Research Fund of Hunan Provincial Education Department (09B007), by the National Natural Science Foundation for Distinguished Young Scholars of China under Grant 40625013, and by the 863 Program under Grant 2007AA12Z311. D. Wen is with the School of Traffic and Transportation Engineering, Changsha University of Science and Technology, Changsha 410004, China, and also with the Institute of Geodesy and Geophysics, Chinese Academy of Sciences, Wuhan 430077, China. Y. Yuan and J. Ou are with the Key Laboratory of Dynamic Geodesy, Institute of Geodesy and Geophysics, Chinese Academy of Sciences, Wuhan 430077, China. K. Zhang is with the School of Mathematical and Geospatial Sciences, Royal Melbourne Institute of Technology University, Melbourne, VIC 3001, Australia. Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/TGRS.2010.2044579
necessary to investigate the ionospheric response to geomagnetic storms using the observations which are near continuous in both temporal and spatial sense. The wide usage and rapid development of GNSS technology has opened a new avenue for the study of the ionosphere. The GNSS-based computerized ionospheric tomography (CIT) technique has recently attracted much attention to overcome the aforementioned problems based on traditional methods. Much research work [2]– [4] has been carried out since its feasibility was first verified by [5]. Recent research has demonstrated that GNSS-based CIT technique can be effectively used to investigate the impact of ionospheric disturbance in a large scale during geomagnetic storms [6]–[11]. In this paper, dual-frequency GNSS observations from the Crustal Movement Observation Network of China (CMONOC) are first used to investigate the ionospheric response to the geomagnetic storm on August 21, 2003, and the dynamics of the ionospheric variations during the geomagnetic storm has been examined. Ionospheric storm phase effects and some key features of the ionospheric structure over China are shown in the tomographic images. II. GNSS DATA P ROCESSING A. Selected GNSS Observations In this paper, the GNSS observation data of CMONOC is used to reconstruct the ionospheric electron density (IED) distribution. The GNSS observation stations are unevenly distributed between the reconstructed region. The sample interval of the used GNSS data is 30 s. Fig. 1 shows the distribution of the used GNSS stations. B. Cycle-Slip Detection and Correction of GNSS Data Routine GNSS data preprocessing is first performed to detect, remove, and/or correct outliers and cycle slips by using the algorithms described by Blewitt [12] prior to computation using the GNSS-based CIT technique. In general, cycle slips are better detectable when dealing with long rather than with short wavelengths. For this reason, the wide-lane phase combination is performed in Blewitt’s algorithm Φδ = Φ1 − Φ2
(1)
where Φ1 and Φ2 are the actually recorded phase of the carriers. The wide-lane wavelength is obviously longer than the GNSS
0196-2892/$26.00 © 2010 IEEE Authorized licensed use limited to: CHANGSHA UNIV OF SCIENCE AND TECHNOLOGY. Downloaded on July 19,2010 at 03:23:30 UTC from IEEE Xplore. Restrictions apply.
WEN et al.: IONOSPHERIC RESPONSE TO GEOMAGNETIC STORM OVER CHINA
3213
derivation of STEC is commonly made using the pseudoranges and phase measurements of the dual-frequency GNSS signals. Since the GNSS phase measurements are less affected by multipath effects, the accuracy of the STEC derived from differential phases (STECph ) is higher than that of the STEC derived from differential pseudoranges (STECco ). Because of the ambiguity in GNSS phase measurements, STECph only provides the relative changes of ionospheric TEC. On the other hand, although P-code pseudoranges are sensitive to multipath effects, (STECco ) values are absolute. Taking into account these facts, an absolute STEC may be obtained by introducing an extra term BL . It can be formulated as follows [13]–[15]: STEC = STECph + BL . Fig. 1.
(7)
Distribution sketch of the used GNSS stations.
wavelengths, and with the corresponding wide-lane ambiguity, Nδ is the difference of the carrier phase ambiguity at both frequencies, i.e., Nδ = N1 − N 2 =
Lδ − Pδ λδ
(2)
where Lδ and Pδ are defined as Lδ =
f1 L1 − f2 L2 f1 − f2
(3)
Pδ =
f1 P1 + f2 P2 . f1 + f2
(4)
During the data analysis, the so-called phase-connected arcs are built. One phase-connected arc consists of a series of data points recorded during a satellite track without detected cycle slip. The detection criterion for a cycle slip is given by the root mean square of the Nδ from one arc, and Nδ is related to the datum of index i with the arc. To determine the connected arcs, the arc-related mean Nδ is iteratively calculated by increasing the number of i successive measurements, as well as its mean square σ 2 Nδi+1 − Nδ i i+1 2 Nδi+1 − Nδ 2 − σi2 2 . = σi + i+1
Nδ i+1 = Nδ i + 2 σi+1
(5)
(6)
The cycle-slip criterion is fulfilled when the following Nδi+1 and Nδi+2 differ by more than 2σi from the actual mean value of the arc Nδ i . Then, a new arc begins to build. If only Nδi+1 and not Nδi+2 deviates by more than 2σi , Nδi+1 is defined as an outlier. In this case, Nδi+1 is deleted and replaced by the simple relation Nδi+1 = (1/2)(Nδi + Nδi+2 ). At each beginning of a new arc, the root mean square is initialized with σ0 = 0.5. III. M ETHOD A. Derivation of STEC To reconstruct the ionospheric IED, the computation of slant total electron content (TEC) (STEC) is necessary. The
If N measurements are obtained during a satellite pass, BL can be modeled by the following equation: N 2 STECcoi − STECphi BL = N. (8) i=1
Due to the instrumental bias of GNSS pseudorange measurements, the derived STECS should be precalibrated. This paper used a strategy to cope with the aforementioned biases. First, the satellite and receiver instrument biases over the three days (20th, 21st, and 22nd of August 2003) were determined using the method described by Yuan and Ou [16], [17]. Mean values of the instrumental biases over the three days were then computed. They were then used to calibrate each of the differential delays for each satellite–receiver pair during the August 21, 2003 storm. Second, the bias-corrected differential delays are applied to calibrate the differential phases of the storm day by using a least squares fitting method. Absolute STECs are determined from the differential phase and time delays recorded during the storm using this procedure. It is well known that pseudoranges from low-elevation-angle satellites are prone to multipath effects. In general, the lower the elevation angle, the larger the multipath error. On the other hand, GNSS data obtained at low elevation angles are important for getting some information about the vertical distribution of the IED. Hence, a critical problem is how to select a proper cutoff elevation angle for the CIT technique. A large cutoff elevation angle reduces the vertical resolution of the inverted results, and a small elevation angle would seriously reduce the accuracy of the IED inversion. Given the aforementioned reasons, an elevation cutoff angle of 10◦ is adopted in this paper. B. Tomographic Theory Ionospheric STEC is the line integral of IED along the ray path from a satellite to a receiver, and it can be defined as STEC = N e(s)ds (9) l
where Ne (s) represents the IED, and l is the ray propagation path of each GNSS satellite-receiver pair.
Authorized licensed use limited to: CHANGSHA UNIV OF SCIENCE AND TECHNOLOGY. Downloaded on July 19,2010 at 03:23:30 UTC from IEEE Xplore. Restrictions apply.
3214
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, VOL. 48, NO. 8, AUGUST 2010
Fig. 2. Two-dimensional image of IED over China at 15:00 UT on [(a) and (c)] geomagnetic quiet days and (b) storm day. (a) August 20, 2003. (b) August 21, 2003. (c) August 22, 2003. The unit of IED is 1011 e/m3 .
From (9), it can be seen that the relation between ionospheric STEC and IED is not linear. To simplify the IED inversion, an imaged region of the ionosphere is first discretized into some small pixels in a selected reference frame. In this paper, the study area of the ionospheric image generation covers 20◦ in latitude (20◦ N–40◦ N) and 20◦ in longitude (100◦ E–120◦ E). The altitude ranges are from 100 to 1000 km. To discretize the reconstruction area, horizontal resolution of 0.5◦ (in latitude) and 1◦ (in longitude) and a vertical resolution of 30 km are used. Within each pixel, the electron density can be assumed constant. The STEC along the ray path can then be expressed as a finite sum of the shortest integrals in all segments of the ray path. Each set of TEC values along the ray path from a satellite to a receiver can be written as STECi =
n
Aij xj + ei .
(10)
j=1
Equation (10) can be generally written in a simple matrix notation as ym×1 = Am×n xn×1 + em×1
(11)
where n is the number of pixels in the image, m is the number of STEC measurements, y is a column vector of the m known STEC measurements, A is an m × n matrix with Aij being the length of ray i traversing pixel j, x is a column vector consisting of all the unknown electron densities in all the pixels, and e is a column vector associated with the discretization errors and measurement noises.
C. Tomographic Algorithm In (11), matrix A is usually rank deficient. It means that only a subset of the m equations is linearly independent. Moreover, the number of independent equations is generally less than the number of the unknown parameters. Therefore, the tomographic reconstruction of the IED is an ill-posed inverse problem due to the restricted view window of the tomographic system and the irregularity and sparsity of the ground-based GNSS stations. To resolve this problem, a hybrid reconstruction algorithm (HRA) is used in this paper. In this algorithm, the truncated singular value decomposition (TSVD) method is first
used to get an approximate solution of the tomographic system; the TSVD of Ak is defined as [18] Ak =
k
Ui Di ViT
(12)
i=1
where k < n. It neglects n − k smallest singular values. Then, the pseudoinverse A+ k is written as A+ k =
k
Vi Di−1 UiT .
(13)
i=1
The solution by TSVD is expressed as xk = A+ k y.
(14)
The estimates obtained from (14) are then input as initial states required by the algebraic reconstruction technique (ART) x(0) = xk .
(15)
Next, the reconstruction is performed with the following ART algorithm:
(16) x(k+1) = x(k) + λk yi − ai x(k) . The column vector λk consisting of all relaxation parameters is given as ai aT (17) λk = γ · aT i i where ai represents the ith row vector in projection matrix A, and γ, called relax parameter, is fixed in ionospheric tomography. In this paper, γ = 1. IV. R ESULTS AND A NALYSIS A. Analysis of the Reconstructed Results A moderate geomagnetic storm event occurred on August 21, 2003. As a result, an initial phase of the geomagnetic storm with a sudden commencement was seen after 03:40 UT. The main phase of the storm lasted from 06:00 UT to 24:00 UT and had a minimum disturbance storm time (Dst) index of −82 nT. The 3-hour Kp index reached 6 at 09:00 UT. Fig. 2 shows the IED variations with altitude and latitude along longitude meridian of 114.5◦ at 15:00 UT (23:00 BT; BT
Authorized licensed use limited to: CHANGSHA UNIV OF SCIENCE AND TECHNOLOGY. Downloaded on July 19,2010 at 03:23:30 UTC from IEEE Xplore. Restrictions apply.
WEN et al.: IONOSPHERIC RESPONSE TO GEOMAGNETIC STORM OVER CHINA
3215
Fig. 3. Three schematic comparisons between the profiles obtained from CIT and those from ionosonde observations at Wuhan at 14:30 UT–15:00 UT (22:30 LT–23:00 LT) on August 20, 21, and 22, 2003. (a) August 20, 2003. (b) August 21, 2003. (c) August 22, 2003. The unit of IED is 1011 e/m3 .
Fig. 4. Variations of IED at longitude of 114.5◦ E for the three time periods on August 21, 2003. (a) 05:30 UT–06:00 UT (13:30 LT–14:00 LT). (b) 08:30 UT–09:00 UT (16:30 LT–17:00 LT). (c) 11:30 UT–12:00 UT (19:30 LT–20:00 LT). The unit of IED is 1011 e/m3 .
is the abbreviation of Beijing time) during the geomagnetically active period on August 21, 2003 and geomagnetically quiet periods on the 20th and 22nd of August 2003, respectively. Comparing Fig. 2(b) with Fig. 2(a), one can see that the IED in the F region was depleted during the geomagnetic storm on August 21, 2003, which appeared to be the negative storm phase effects. The maximum depletion of the IED reached 30%. The positive storm phase effects of the ionosphere however appeared latitudinally from 20◦ N to 37◦ N and altitudinally from 500 to 700 km above the ground, and then the negative storm phase effects also appeared latitudinally from latitude 37◦ N to latitude 40◦ N. The reconstructed results show that the storm phase effects are not only latitude dependent but also altitude dependent. Similar characteristics of the geomagnetic storm can also be found in other periods. Fig. 2(b) shows an elongated region of reduced electron density at a latitude around 32◦ N. The IED value of the trough is about 5.7 × 1011 e/m3 . In the north of the ionospheric trough, the IED increases and reaches a peak at 36◦ N where the peak height is about 320 km; the peak density of the ionosphere increases to 6.2 × 1011 e/m3 , and then the IED begins to decrease. However, the peak IED reaches 8.0 × 1011 e/m3 south of the trough, and the peak height is about 350 km, which is higher than the north peak height of the ionospheric trough. At the same time, an intensively disturbed structure of the ionosphere appears between 20◦ N and 35◦ N; this demonstrates that the GNSS-based CIT technique is a powerful tool to monitor and investigate largescale ionospheric structure under disturbed conditions. From Fig. 2(c), it can be seen that the IED gradient gradually comes back to the state before the storm occurred.
The available ionosonde station at Wuhan in China region provided independent comparisons with the tomographically obtained electron-density profiles. Fig. 3 provided verification of the outputs of CIT, for 14:30 UT–15:00 UT (22:30 local time (LT)–23:00 LT) time period on August 20, 21, and 22, 2003, with the available valid ionosonde data recorded at Wuhan station. From Fig. 3, the agreement between the profiles obtained from tomographic reconstruction and those from ionosonde data can be seen. This validated the reliability of the tomographic results. The previous tomographic results show only the nighttime density profiles. To further examine the ionospheric response to the storm event in daytime and dusk, Fig. 4 shows a series of typical tomographic images of the ionosphere over China along a longitude chain of 114.5◦ E for the 06:00 UT–12:00 UT period during the storm on August 21, 2003. The label in the top-right corner represents the universal time. From Fig. 4, one can see that the disturbed structure of the ionosphere shown in Fig. 3(b) can be clearly seen, and the dynamics of the ionospheric response to the geomagnetic storm is also shown. However, the IED gradient in Fig. 4(a) obtained at 05:30 UT–06:00 UT (13:30 LT–14:00 LT) is smaller than those in Fig. 4(b) and (c). The ranges of the disturbed ionospheric structure widened with time. This characteristic of the disturbed structure is also seen in the reconstructed image in nighttime density profile [in Fig. 3(b)]. To validate the reliability of the reconstructed results shown in Fig. 4, three comparisons are made in Fig. 5. Fig. 5 shows that the profiles obtained from the CIT at three time periods (05:30 UT–06:00 UT, 08:30 UT–09:00 UT, and 11:30 UT–12:00 UT)
Authorized licensed use limited to: CHANGSHA UNIV OF SCIENCE AND TECHNOLOGY. Downloaded on July 19,2010 at 03:23:30 UTC from IEEE Xplore. Restrictions apply.
3216
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, VOL. 48, NO. 8, AUGUST 2010
Fig. 5. Three schematic comparisons between the profiles obtained from CIT and those from ionosonde observations at Wuhan station at three time periods. (a) 05:30 UT–06:00 UT (13:30 LT–14:00 LT). (b) 08:30 UT–09:00 UT (16:30 LT–17:00 LT). (c) 11:30 UT–12:00 UT (19:30 LT–20:00 LT). The unit of IED is 1011 e/m3 . TABLE I S TATISTICS OF THE R ECONSTRUCTED IED E RROR U SING ART, TSVD, AND HRA
agreed with those from ionosonde data recorded at Wuhan station. The comparison verified the validation of the tomographic reconstruction in Fig. 4. B. Statistics of the Reconstructed Errors To show the superiority of the aforementioned HRA, the statistics of the reconstructed error in the tomographic inversion is made. Table I gives the statistics results. From Table I, we can see that the maximum of the reconstructed IED error is 1.24 × 1010 e/m3 , the minimum of the reconstructed IED error is −1.16 × 1010 e/m3 , and the average of the reconstructed IED error is 1.5 × 1010 e/m3 , which are very small compared with the typical peak density of 1.78 × 1012 el/m3 . By comparison with the TSVD and ART alone, the reconstructed results of HRA are obviously improved. V. D ISCUSSION AND C ONCLUSION A GNSS-based CIT technique has been firstly applied to investigate the ionospheric response to the geomagnetic storm on August 21, 2003 over China. The reconstructed images of the IED show that a large-scale ionospheric disturbance is evident in China with ionization depletion in the F region and ionization enhancement above the F region, which represent a negative and a positive storm phase effect, respectively. Similar to early theoretical analyses by Buonsanto and the references therein, the negative storm phases that occurred in the F region during the geomagnetic storm on August 21, 2003 could be attributed to an increased input of energy to the region and an expan-
sion of the neutral atmosphere. Such a rapid expansion may cause upwelling, which in turn induces a dramatic depletion of the atomic to molecular neutral density ratio. This change in chemical composition may cause an increased recombination in the ionosphere and a reduction in ionization concentration. The positive phases are generally believed to be caused by uplifting of the F region by equatorward winds in the early stage of a storm development. The reconstructed results suggest that one may monitor and investigate the disturbed variations of the IED in China during geomagnetic storm with the CIT method and dual-frequency GNSS data from CMONOC. This is helpful in understanding finer structures of the ionosphere during geomagnetic storms. The tomographic results have illustrated that the intense disturbance of the ionosphere in China occurred during the storm, and the ionospheric storm phase effects are not only latitude dependent but also altitude dependent. We are convinced that these results are beneficial to the understanding of the characteristics and variation mechanism of the ionosphere in China during geomagnetic storms, and they are capable of providing a valuable experimental support for understanding the complex behavior of the F region during geomagnetic storms. R EFERENCES [1] P. Yin, C. Mitchell, and G. Bust, “Observation of the F region height redistribution in the storm-time ionosphere over Europe and the USA using GNSS imaging,” Geophys. Res. Lett., vol. 33, no. 18, p. L18 803, Sep. 2006. DOI: 10.1029/2006GL027125. [2] A. K. Shukla, S. Das, N. Nagori, M. R. Sivaraman, and K. Bandyopadhyay, “Two-shell ionospheric model for Indian region: A novel approach,” IEEE Trans. Geosci. Remote Sens., vol. 47, no. 8, pp. 1407–2412, Aug. 2009.
Authorized licensed use limited to: CHANGSHA UNIV OF SCIENCE AND TECHNOLOGY. Downloaded on July 19,2010 at 03:23:30 UTC from IEEE Xplore. Restrictions apply.
WEN et al.: IONOSPHERIC RESPONSE TO GEOMAGNETIC STORM OVER CHINA
[3] C. L. Mai and J. F. King, “Reconstruction of ionospheric perturbation induced by 2004 Sumatra tsunami using a computerized tomography technique,” IEEE Trans. Geosci. Remote Sens., vol. 47, no. 10, pp. 3303– 3312, Oct. 2009. [4] J. K. Lee and F. Kamalabadi, “GNSS-based radio tomography with edgepreserving regularization,” IEEE Trans. Geosci. Remote Sens., vol. 47, no. 1, pp. 312–324, Jan. 2009. [5] V. Kunitsyn, E. Andreeva, and O. Razinkov, “Possibilities of the nearspace environment radio tomography,” Radio Sci., vol. 32, no. 5, pp. 1953–1963, 1997. [6] M. Hernandez-Pajares, J. M. Juan, and J. Sanz, “Global observation of the ionosphere electronic response to solar events using ground and LEO GNSS data,” J. Geophys. Res., vol. 103, no. A9, pp. 20 789–20 796, 1998. [7] M. J. Buonsanto, “Ionospheric storms—A review,” Space Sci. Rev., vol. 88, no. 3/4, pp. 563–601, Apr. 1999. [8] G. S. Bust, D. Coco, and J. J. Makela, “Combined ionospheric campaign 1: Ionospheric tomography and GNSS total electron content (TEC) depletions,” Geophys. Res. Lett., vol. 27, no. 18, pp. 2849–2852, 2000. [9] E. Yizengaw, E. A. Essex, and R. Birsa, “The southern hemisphere and equatorial region ionization response for a 22 September 1999 severe magnetic storm,” Ann. Geophys., vol. 22, no. 8, pp. 2765–2773, Aug. 2004. [10] E. Yizengaw, P. L. Dyson, E. A. Essex, and M. B. Moldwin, “Ionosphere dynamics over the Southern Hemisphere during the 31 March 2001 severe magnetic storm using multi-instrument measurement data,” Ann. Geophys., vol. 23, no. 3, pp. 707–721, Mar. 2005. [11] L. R. Cander and S. J. Mihajlovic, “Ionospheric spatial and temporal variations during the 29–31 October 2003 storm,” J. Atmos. Sol. Terr. Phys., vol. 67, no. 12, pp. 1118–1128, Aug. 2005. [12] G. Blewitt, “An automatic editing algorithm for GNSS data,” Geophys. Res. Lett., vol. 17, no. 3, pp. 199–202, 1990. [13] D. Wen, Y. Yuan, and J. Ou, “Monitoring the three-dimensional ionospheric electron density distribution using GNSS data over China,” J. Earth Syst. Sci., vol. 116, no. 3, pp. 235–244, Jun. 2007. [14] D. Wen, Y. Yuan, and J. Ou, “Three-dimensional ionospheric tomography by an improved algebraic reconstruction technique,” GPS Solutions, vol. 11, no. 4, pp. 253–258, Nov. 2007. [15] Y. Yuan, D. Wen, J. Ou, X. Huo, R. Yang, K. Zhang, and R. Grenfel, “Preliminary research on imaging the ionosphere using CIT and China permanent GNSS tracking station data,” in Proc. Int. Assoc. Geodesy Symposia, P. Tregoning and C. Rizos, Eds., Cairns, Australia, Aug. 22–25, 2005, vol. 130, pp. 876–883. [16] Y. Yuan and J. Ou, “The effects of instrumental bias in GNSS observations on determining ionospheric delays and the methods of its calibration,” Acta Geodaetica et Cartographica Sinica, vol. 28, pp. 110–114, 1999. [17] Y. Yuan, X. Huo, and J. Ou, “Models and methods for precise determination of ionospheric delays using GNSS,” Prog. Nat. Sci., vol. 17, no. 2, pp. 187–196, 2007. [18] D. Wen, Y. Yuan, J. Ou, K. Zhang, and K. Liu, “A hybrid reconstruction algorithm for 3-D ionospheric tomography,” IEEE Trans. Geosci. Remote Sens., vol. 46, no. 6, pp. 1733–1739, Jun. 2008.
Debao Wen received the B.Sc. degree in surveying engineering from Henan Polytechnic University, Jiaozuo, China, and the Ph.D. degree in geodesy and surveying engineering from the Chinese Academy of Sciences (CAS), Wuhan, China, in 2008. He is currently with the Changsha University of Science and Technology, Changsha, China, and with the Institute of Geodesy and Geophysics, CAS, Wuhan, China. His current research projects include 3-D ionospheric/atmospheric tomography based on GNSS observations and the effects of space weather on navigation and positioning. Dr. Wen was the recipient of several awards, including the Second Award for National Science and Technology Progress of China, the Second Award of the Chinese Military Science and Technology Advance, the Excellent Prize of the Presidential Scholarship of CAS, and the Student Best Paper Award of ION GNSS 2007.
3217
Yunbin Yuan received the B.Sc. degree in surveying engineering from the Shandong University of Science and Technology, Qingdao, China, in 1995 and the Ph.D. degree in geodesy and surveying from the Institute of Geodesy and Geophysics, Chinese Academy of Sciences (CAS), Beijing, China, in 2002. He is currently with Institute of Geodesy and Geophysics, CAS. His main subjects include satellite geodesy and GNSS positioning/atmosphere/ ionosphere. His current research interests are GNSSbased spatial environmental monitoring and analysis, including the ionospheric correction for Network RTK for centimeter-level realtime positioning; high-precision GNSS satellite navigation and positioning in different augmentation systems such as WAAS and Network RTK; GNSS in applications to orbit determination for Low Earth orbiter satellites; and China’s manned space engineering. Dr. Yuan was the recipient of several awards, including the Prize of Chinese Military Science and Technology Advance, Prize of Significant Contributions to China’s Manned Spaceflight, China’s Best Ph.D. thesis, Special Prize of the President Scholarship (CAS), and the 2001 ION GPS Student Best Paper Award.
Jikun Ou received the B.Sc. degree in mechanical engineering from the Beijing Machine College of China, Beijing, China, in 1970 and the M.Sc. and Ph.D. degrees in geodesy from the Institute of Geodesy and Geophysics, (IGG) Chinese Academy of Sciences (CAS), Beijing, in 1982 and 1994, respectively. He is currently a Research Professor with IGG, and a Doctoral Supervisor of the Graduate University, CAS. He is an Editor of five journals. Since 1980, he has been performing research on the theory of errors in geodesy and surveying. He is the author of more than 100 scientific and technical research articles. He has obtained a special allowance from the State Council of China since 1993, and was chosen as a specialist of outstanding contributions in Hubei province. His current major areas of expertise include the control of data quality, detection of gross errors, precise positioning by GPS, atmosphere and its effects on GPS surveying, and the approach of precise orbit determination for low orbital satellites and GNSS satellites, and so on.
Kefei Zhang received the B.Sc. and M.Sc. degrees in geodesy from Wuhan University (formerly Wuhan Technical University of Surveying and Mapping), Wuhan, China, in 1985 and 1988, respectively, and the Ph.D. degree in geodesy from the Curtin University of Technology, Perth, Australia, in 1997. Prior to joining the Royal Melbourne Institute of Technology (RMIT) University, Melbourne, Australia, in 1999, he was with the Institute of Engineering Surveying and Space Geodesy, University of Nottingham, Nottingham, U.K., as a Postdoctoral Researcher for about three years. He is currently an Associate Professor and leads the Measurement Science Group with the School of Mathematical and Geospatial Sciences, RMIT. He has a mixed background of satellite positioning, tracking, and geodesy. He is also a Guest Professor with a few international universities. His current research projects primarily involve algorithm development and innovative applications of GPS and GNSS technologies for high-accuracy positioning, atmospheric and meteorological studies, network-based RTK, and people and object tracking. In addition, he is also interested in geodesy-related theoretical and practical studies such as Earth’s gravity field modeling, satelliteorbit determination, and spectral and fractal analyses. He has been the author of about 100 peer-reviewed publications in these fields since 1990. Dr. Zhang is a member of a number of international special study groups and editorial boards of scientific journals. He is the President of the International Association of Chinese Professionals in Global Positioning Systems.
Authorized licensed use limited to: CHANGSHA UNIV OF SCIENCE AND TECHNOLOGY. Downloaded on July 19,2010 at 03:23:30 UTC from IEEE Xplore. Restrictions apply.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, VOL. 48, NO. 8, AUGUST 2010
Ionospheric Response to the Geomagnetic Storm on August 21, 2003 Over China Using GNSS-Based Tomographic Technique Debao Wen, Yunbin Yuan, Jikun Ou, and Kefei Zhang
Abstract—The impacts of the August 21, 2003 geomagnetic storm on the ionosphere over China have been first investigated by using the so-called computerized ionospheric tomography (CIT) technique and the observations of the Crustal Movement Observation Network of China. Tomographic results show that the main ionospheric effects of this geomagnetic storm over China are as follows: 1) the negative storm phase effect appears in the F region and 2) the positive storm phase effect occurs above the F region. Meanwhile, some key features in the ionospheric structure have been revealed in the ionospheric images during the storm; this includes the disturbances and an elongated region of the reduced electron density at the latitude around 32◦ N. Statistical comparisons are carried out to confirm the reliability of the global-navigation-satellite-system-based CIT reconstruction results using the profile obtained from ionosonde observations. Index Terms—Electron density, geomagnetic storm, global navigation satellite systems (GNSS), ionosphere, tomography.
I. I NTRODUCTION
T
HE OCCURRENCE of geomagnetic storms greatly changes the spatial distribution of ionization. This may also affect the propagation of radio signals and deteriorate the performance of global navigation satellite systems (GNSS). In the past, traditional ground-based instruments, such as incoherent scatter radars (ISRs) and ionosondes, have been extensively used for the investigation of the effects of geomagnetic storms on the ionosphere. However, ionosondes cannot be used to measure the topside ionosphere, and sometimes they suffer from absorption problems during geomagnetic storms, while ISRs are limited by geographical locations [1]. Given these limitations and the global impacts of geomagnetic storm, it is Manuscript received July 14, 2009; revised December 15, 2009. Date of publication April 8, 2010; date of current version July 21, 2010. This work was supported in part by the National Science Foundation of China under Grants 40804002 and 40890160, by the Scientific Research Fund of Hunan Provincial Education Department (09B007), by the National Natural Science Foundation for Distinguished Young Scholars of China under Grant 40625013, and by the 863 Program under Grant 2007AA12Z311. D. Wen is with the School of Traffic and Transportation Engineering, Changsha University of Science and Technology, Changsha 410004, China, and also with the Institute of Geodesy and Geophysics, Chinese Academy of Sciences, Wuhan 430077, China. Y. Yuan and J. Ou are with the Key Laboratory of Dynamic Geodesy, Institute of Geodesy and Geophysics, Chinese Academy of Sciences, Wuhan 430077, China. K. Zhang is with the School of Mathematical and Geospatial Sciences, Royal Melbourne Institute of Technology University, Melbourne, VIC 3001, Australia. Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/TGRS.2010.2044579
necessary to investigate the ionospheric response to geomagnetic storms using the observations which are near continuous in both temporal and spatial sense. The wide usage and rapid development of GNSS technology has opened a new avenue for the study of the ionosphere. The GNSS-based computerized ionospheric tomography (CIT) technique has recently attracted much attention to overcome the aforementioned problems based on traditional methods. Much research work [2]– [4] has been carried out since its feasibility was first verified by [5]. Recent research has demonstrated that GNSS-based CIT technique can be effectively used to investigate the impact of ionospheric disturbance in a large scale during geomagnetic storms [6]–[11]. In this paper, dual-frequency GNSS observations from the Crustal Movement Observation Network of China (CMONOC) are first used to investigate the ionospheric response to the geomagnetic storm on August 21, 2003, and the dynamics of the ionospheric variations during the geomagnetic storm has been examined. Ionospheric storm phase effects and some key features of the ionospheric structure over China are shown in the tomographic images. II. GNSS DATA P ROCESSING A. Selected GNSS Observations In this paper, the GNSS observation data of CMONOC is used to reconstruct the ionospheric electron density (IED) distribution. The GNSS observation stations are unevenly distributed between the reconstructed region. The sample interval of the used GNSS data is 30 s. Fig. 1 shows the distribution of the used GNSS stations. B. Cycle-Slip Detection and Correction of GNSS Data Routine GNSS data preprocessing is first performed to detect, remove, and/or correct outliers and cycle slips by using the algorithms described by Blewitt [12] prior to computation using the GNSS-based CIT technique. In general, cycle slips are better detectable when dealing with long rather than with short wavelengths. For this reason, the wide-lane phase combination is performed in Blewitt’s algorithm Φδ = Φ1 − Φ2
(1)
where Φ1 and Φ2 are the actually recorded phase of the carriers. The wide-lane wavelength is obviously longer than the GNSS
0196-2892/$26.00 © 2010 IEEE Authorized licensed use limited to: CHANGSHA UNIV OF SCIENCE AND TECHNOLOGY. Downloaded on July 19,2010 at 03:23:30 UTC from IEEE Xplore. Restrictions apply.
WEN et al.: IONOSPHERIC RESPONSE TO GEOMAGNETIC STORM OVER CHINA
3213
derivation of STEC is commonly made using the pseudoranges and phase measurements of the dual-frequency GNSS signals. Since the GNSS phase measurements are less affected by multipath effects, the accuracy of the STEC derived from differential phases (STECph ) is higher than that of the STEC derived from differential pseudoranges (STECco ). Because of the ambiguity in GNSS phase measurements, STECph only provides the relative changes of ionospheric TEC. On the other hand, although P-code pseudoranges are sensitive to multipath effects, (STECco ) values are absolute. Taking into account these facts, an absolute STEC may be obtained by introducing an extra term BL . It can be formulated as follows [13]–[15]: STEC = STECph + BL . Fig. 1.
(7)
Distribution sketch of the used GNSS stations.
wavelengths, and with the corresponding wide-lane ambiguity, Nδ is the difference of the carrier phase ambiguity at both frequencies, i.e., Nδ = N1 − N 2 =
Lδ − Pδ λδ
(2)
where Lδ and Pδ are defined as Lδ =
f1 L1 − f2 L2 f1 − f2
(3)
Pδ =
f1 P1 + f2 P2 . f1 + f2
(4)
During the data analysis, the so-called phase-connected arcs are built. One phase-connected arc consists of a series of data points recorded during a satellite track without detected cycle slip. The detection criterion for a cycle slip is given by the root mean square of the Nδ from one arc, and Nδ is related to the datum of index i with the arc. To determine the connected arcs, the arc-related mean Nδ is iteratively calculated by increasing the number of i successive measurements, as well as its mean square σ 2 Nδi+1 − Nδ i i+1 2 Nδi+1 − Nδ 2 − σi2 2 . = σi + i+1
Nδ i+1 = Nδ i + 2 σi+1
(5)
(6)
The cycle-slip criterion is fulfilled when the following Nδi+1 and Nδi+2 differ by more than 2σi from the actual mean value of the arc Nδ i . Then, a new arc begins to build. If only Nδi+1 and not Nδi+2 deviates by more than 2σi , Nδi+1 is defined as an outlier. In this case, Nδi+1 is deleted and replaced by the simple relation Nδi+1 = (1/2)(Nδi + Nδi+2 ). At each beginning of a new arc, the root mean square is initialized with σ0 = 0.5. III. M ETHOD A. Derivation of STEC To reconstruct the ionospheric IED, the computation of slant total electron content (TEC) (STEC) is necessary. The
If N measurements are obtained during a satellite pass, BL can be modeled by the following equation: N 2 STECcoi − STECphi BL = N. (8) i=1
Due to the instrumental bias of GNSS pseudorange measurements, the derived STECS should be precalibrated. This paper used a strategy to cope with the aforementioned biases. First, the satellite and receiver instrument biases over the three days (20th, 21st, and 22nd of August 2003) were determined using the method described by Yuan and Ou [16], [17]. Mean values of the instrumental biases over the three days were then computed. They were then used to calibrate each of the differential delays for each satellite–receiver pair during the August 21, 2003 storm. Second, the bias-corrected differential delays are applied to calibrate the differential phases of the storm day by using a least squares fitting method. Absolute STECs are determined from the differential phase and time delays recorded during the storm using this procedure. It is well known that pseudoranges from low-elevation-angle satellites are prone to multipath effects. In general, the lower the elevation angle, the larger the multipath error. On the other hand, GNSS data obtained at low elevation angles are important for getting some information about the vertical distribution of the IED. Hence, a critical problem is how to select a proper cutoff elevation angle for the CIT technique. A large cutoff elevation angle reduces the vertical resolution of the inverted results, and a small elevation angle would seriously reduce the accuracy of the IED inversion. Given the aforementioned reasons, an elevation cutoff angle of 10◦ is adopted in this paper. B. Tomographic Theory Ionospheric STEC is the line integral of IED along the ray path from a satellite to a receiver, and it can be defined as STEC = N e(s)ds (9) l
where Ne (s) represents the IED, and l is the ray propagation path of each GNSS satellite-receiver pair.
Authorized licensed use limited to: CHANGSHA UNIV OF SCIENCE AND TECHNOLOGY. Downloaded on July 19,2010 at 03:23:30 UTC from IEEE Xplore. Restrictions apply.
3214
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, VOL. 48, NO. 8, AUGUST 2010
Fig. 2. Two-dimensional image of IED over China at 15:00 UT on [(a) and (c)] geomagnetic quiet days and (b) storm day. (a) August 20, 2003. (b) August 21, 2003. (c) August 22, 2003. The unit of IED is 1011 e/m3 .
From (9), it can be seen that the relation between ionospheric STEC and IED is not linear. To simplify the IED inversion, an imaged region of the ionosphere is first discretized into some small pixels in a selected reference frame. In this paper, the study area of the ionospheric image generation covers 20◦ in latitude (20◦ N–40◦ N) and 20◦ in longitude (100◦ E–120◦ E). The altitude ranges are from 100 to 1000 km. To discretize the reconstruction area, horizontal resolution of 0.5◦ (in latitude) and 1◦ (in longitude) and a vertical resolution of 30 km are used. Within each pixel, the electron density can be assumed constant. The STEC along the ray path can then be expressed as a finite sum of the shortest integrals in all segments of the ray path. Each set of TEC values along the ray path from a satellite to a receiver can be written as STECi =
n
Aij xj + ei .
(10)
j=1
Equation (10) can be generally written in a simple matrix notation as ym×1 = Am×n xn×1 + em×1
(11)
where n is the number of pixels in the image, m is the number of STEC measurements, y is a column vector of the m known STEC measurements, A is an m × n matrix with Aij being the length of ray i traversing pixel j, x is a column vector consisting of all the unknown electron densities in all the pixels, and e is a column vector associated with the discretization errors and measurement noises.
C. Tomographic Algorithm In (11), matrix A is usually rank deficient. It means that only a subset of the m equations is linearly independent. Moreover, the number of independent equations is generally less than the number of the unknown parameters. Therefore, the tomographic reconstruction of the IED is an ill-posed inverse problem due to the restricted view window of the tomographic system and the irregularity and sparsity of the ground-based GNSS stations. To resolve this problem, a hybrid reconstruction algorithm (HRA) is used in this paper. In this algorithm, the truncated singular value decomposition (TSVD) method is first
used to get an approximate solution of the tomographic system; the TSVD of Ak is defined as [18] Ak =
k
Ui Di ViT
(12)
i=1
where k < n. It neglects n − k smallest singular values. Then, the pseudoinverse A+ k is written as A+ k =
k
Vi Di−1 UiT .
(13)
i=1
The solution by TSVD is expressed as xk = A+ k y.
(14)
The estimates obtained from (14) are then input as initial states required by the algebraic reconstruction technique (ART) x(0) = xk .
(15)
Next, the reconstruction is performed with the following ART algorithm:
(16) x(k+1) = x(k) + λk yi − ai x(k) . The column vector λk consisting of all relaxation parameters is given as ai aT (17) λk = γ · aT i i where ai represents the ith row vector in projection matrix A, and γ, called relax parameter, is fixed in ionospheric tomography. In this paper, γ = 1. IV. R ESULTS AND A NALYSIS A. Analysis of the Reconstructed Results A moderate geomagnetic storm event occurred on August 21, 2003. As a result, an initial phase of the geomagnetic storm with a sudden commencement was seen after 03:40 UT. The main phase of the storm lasted from 06:00 UT to 24:00 UT and had a minimum disturbance storm time (Dst) index of −82 nT. The 3-hour Kp index reached 6 at 09:00 UT. Fig. 2 shows the IED variations with altitude and latitude along longitude meridian of 114.5◦ at 15:00 UT (23:00 BT; BT
Authorized licensed use limited to: CHANGSHA UNIV OF SCIENCE AND TECHNOLOGY. Downloaded on July 19,2010 at 03:23:30 UTC from IEEE Xplore. Restrictions apply.
WEN et al.: IONOSPHERIC RESPONSE TO GEOMAGNETIC STORM OVER CHINA
3215
Fig. 3. Three schematic comparisons between the profiles obtained from CIT and those from ionosonde observations at Wuhan at 14:30 UT–15:00 UT (22:30 LT–23:00 LT) on August 20, 21, and 22, 2003. (a) August 20, 2003. (b) August 21, 2003. (c) August 22, 2003. The unit of IED is 1011 e/m3 .
Fig. 4. Variations of IED at longitude of 114.5◦ E for the three time periods on August 21, 2003. (a) 05:30 UT–06:00 UT (13:30 LT–14:00 LT). (b) 08:30 UT–09:00 UT (16:30 LT–17:00 LT). (c) 11:30 UT–12:00 UT (19:30 LT–20:00 LT). The unit of IED is 1011 e/m3 .
is the abbreviation of Beijing time) during the geomagnetically active period on August 21, 2003 and geomagnetically quiet periods on the 20th and 22nd of August 2003, respectively. Comparing Fig. 2(b) with Fig. 2(a), one can see that the IED in the F region was depleted during the geomagnetic storm on August 21, 2003, which appeared to be the negative storm phase effects. The maximum depletion of the IED reached 30%. The positive storm phase effects of the ionosphere however appeared latitudinally from 20◦ N to 37◦ N and altitudinally from 500 to 700 km above the ground, and then the negative storm phase effects also appeared latitudinally from latitude 37◦ N to latitude 40◦ N. The reconstructed results show that the storm phase effects are not only latitude dependent but also altitude dependent. Similar characteristics of the geomagnetic storm can also be found in other periods. Fig. 2(b) shows an elongated region of reduced electron density at a latitude around 32◦ N. The IED value of the trough is about 5.7 × 1011 e/m3 . In the north of the ionospheric trough, the IED increases and reaches a peak at 36◦ N where the peak height is about 320 km; the peak density of the ionosphere increases to 6.2 × 1011 e/m3 , and then the IED begins to decrease. However, the peak IED reaches 8.0 × 1011 e/m3 south of the trough, and the peak height is about 350 km, which is higher than the north peak height of the ionospheric trough. At the same time, an intensively disturbed structure of the ionosphere appears between 20◦ N and 35◦ N; this demonstrates that the GNSS-based CIT technique is a powerful tool to monitor and investigate largescale ionospheric structure under disturbed conditions. From Fig. 2(c), it can be seen that the IED gradient gradually comes back to the state before the storm occurred.
The available ionosonde station at Wuhan in China region provided independent comparisons with the tomographically obtained electron-density profiles. Fig. 3 provided verification of the outputs of CIT, for 14:30 UT–15:00 UT (22:30 local time (LT)–23:00 LT) time period on August 20, 21, and 22, 2003, with the available valid ionosonde data recorded at Wuhan station. From Fig. 3, the agreement between the profiles obtained from tomographic reconstruction and those from ionosonde data can be seen. This validated the reliability of the tomographic results. The previous tomographic results show only the nighttime density profiles. To further examine the ionospheric response to the storm event in daytime and dusk, Fig. 4 shows a series of typical tomographic images of the ionosphere over China along a longitude chain of 114.5◦ E for the 06:00 UT–12:00 UT period during the storm on August 21, 2003. The label in the top-right corner represents the universal time. From Fig. 4, one can see that the disturbed structure of the ionosphere shown in Fig. 3(b) can be clearly seen, and the dynamics of the ionospheric response to the geomagnetic storm is also shown. However, the IED gradient in Fig. 4(a) obtained at 05:30 UT–06:00 UT (13:30 LT–14:00 LT) is smaller than those in Fig. 4(b) and (c). The ranges of the disturbed ionospheric structure widened with time. This characteristic of the disturbed structure is also seen in the reconstructed image in nighttime density profile [in Fig. 3(b)]. To validate the reliability of the reconstructed results shown in Fig. 4, three comparisons are made in Fig. 5. Fig. 5 shows that the profiles obtained from the CIT at three time periods (05:30 UT–06:00 UT, 08:30 UT–09:00 UT, and 11:30 UT–12:00 UT)
Authorized licensed use limited to: CHANGSHA UNIV OF SCIENCE AND TECHNOLOGY. Downloaded on July 19,2010 at 03:23:30 UTC from IEEE Xplore. Restrictions apply.
3216
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, VOL. 48, NO. 8, AUGUST 2010
Fig. 5. Three schematic comparisons between the profiles obtained from CIT and those from ionosonde observations at Wuhan station at three time periods. (a) 05:30 UT–06:00 UT (13:30 LT–14:00 LT). (b) 08:30 UT–09:00 UT (16:30 LT–17:00 LT). (c) 11:30 UT–12:00 UT (19:30 LT–20:00 LT). The unit of IED is 1011 e/m3 . TABLE I S TATISTICS OF THE R ECONSTRUCTED IED E RROR U SING ART, TSVD, AND HRA
agreed with those from ionosonde data recorded at Wuhan station. The comparison verified the validation of the tomographic reconstruction in Fig. 4. B. Statistics of the Reconstructed Errors To show the superiority of the aforementioned HRA, the statistics of the reconstructed error in the tomographic inversion is made. Table I gives the statistics results. From Table I, we can see that the maximum of the reconstructed IED error is 1.24 × 1010 e/m3 , the minimum of the reconstructed IED error is −1.16 × 1010 e/m3 , and the average of the reconstructed IED error is 1.5 × 1010 e/m3 , which are very small compared with the typical peak density of 1.78 × 1012 el/m3 . By comparison with the TSVD and ART alone, the reconstructed results of HRA are obviously improved. V. D ISCUSSION AND C ONCLUSION A GNSS-based CIT technique has been firstly applied to investigate the ionospheric response to the geomagnetic storm on August 21, 2003 over China. The reconstructed images of the IED show that a large-scale ionospheric disturbance is evident in China with ionization depletion in the F region and ionization enhancement above the F region, which represent a negative and a positive storm phase effect, respectively. Similar to early theoretical analyses by Buonsanto and the references therein, the negative storm phases that occurred in the F region during the geomagnetic storm on August 21, 2003 could be attributed to an increased input of energy to the region and an expan-
sion of the neutral atmosphere. Such a rapid expansion may cause upwelling, which in turn induces a dramatic depletion of the atomic to molecular neutral density ratio. This change in chemical composition may cause an increased recombination in the ionosphere and a reduction in ionization concentration. The positive phases are generally believed to be caused by uplifting of the F region by equatorward winds in the early stage of a storm development. The reconstructed results suggest that one may monitor and investigate the disturbed variations of the IED in China during geomagnetic storm with the CIT method and dual-frequency GNSS data from CMONOC. This is helpful in understanding finer structures of the ionosphere during geomagnetic storms. The tomographic results have illustrated that the intense disturbance of the ionosphere in China occurred during the storm, and the ionospheric storm phase effects are not only latitude dependent but also altitude dependent. We are convinced that these results are beneficial to the understanding of the characteristics and variation mechanism of the ionosphere in China during geomagnetic storms, and they are capable of providing a valuable experimental support for understanding the complex behavior of the F region during geomagnetic storms. R EFERENCES [1] P. Yin, C. Mitchell, and G. Bust, “Observation of the F region height redistribution in the storm-time ionosphere over Europe and the USA using GNSS imaging,” Geophys. Res. Lett., vol. 33, no. 18, p. L18 803, Sep. 2006. DOI: 10.1029/2006GL027125. [2] A. K. Shukla, S. Das, N. Nagori, M. R. Sivaraman, and K. Bandyopadhyay, “Two-shell ionospheric model for Indian region: A novel approach,” IEEE Trans. Geosci. Remote Sens., vol. 47, no. 8, pp. 1407–2412, Aug. 2009.
Authorized licensed use limited to: CHANGSHA UNIV OF SCIENCE AND TECHNOLOGY. Downloaded on July 19,2010 at 03:23:30 UTC from IEEE Xplore. Restrictions apply.
WEN et al.: IONOSPHERIC RESPONSE TO GEOMAGNETIC STORM OVER CHINA
[3] C. L. Mai and J. F. King, “Reconstruction of ionospheric perturbation induced by 2004 Sumatra tsunami using a computerized tomography technique,” IEEE Trans. Geosci. Remote Sens., vol. 47, no. 10, pp. 3303– 3312, Oct. 2009. [4] J. K. Lee and F. Kamalabadi, “GNSS-based radio tomography with edgepreserving regularization,” IEEE Trans. Geosci. Remote Sens., vol. 47, no. 1, pp. 312–324, Jan. 2009. [5] V. Kunitsyn, E. Andreeva, and O. Razinkov, “Possibilities of the nearspace environment radio tomography,” Radio Sci., vol. 32, no. 5, pp. 1953–1963, 1997. [6] M. Hernandez-Pajares, J. M. Juan, and J. Sanz, “Global observation of the ionosphere electronic response to solar events using ground and LEO GNSS data,” J. Geophys. Res., vol. 103, no. A9, pp. 20 789–20 796, 1998. [7] M. J. Buonsanto, “Ionospheric storms—A review,” Space Sci. Rev., vol. 88, no. 3/4, pp. 563–601, Apr. 1999. [8] G. S. Bust, D. Coco, and J. J. Makela, “Combined ionospheric campaign 1: Ionospheric tomography and GNSS total electron content (TEC) depletions,” Geophys. Res. Lett., vol. 27, no. 18, pp. 2849–2852, 2000. [9] E. Yizengaw, E. A. Essex, and R. Birsa, “The southern hemisphere and equatorial region ionization response for a 22 September 1999 severe magnetic storm,” Ann. Geophys., vol. 22, no. 8, pp. 2765–2773, Aug. 2004. [10] E. Yizengaw, P. L. Dyson, E. A. Essex, and M. B. Moldwin, “Ionosphere dynamics over the Southern Hemisphere during the 31 March 2001 severe magnetic storm using multi-instrument measurement data,” Ann. Geophys., vol. 23, no. 3, pp. 707–721, Mar. 2005. [11] L. R. Cander and S. J. Mihajlovic, “Ionospheric spatial and temporal variations during the 29–31 October 2003 storm,” J. Atmos. Sol. Terr. Phys., vol. 67, no. 12, pp. 1118–1128, Aug. 2005. [12] G. Blewitt, “An automatic editing algorithm for GNSS data,” Geophys. Res. Lett., vol. 17, no. 3, pp. 199–202, 1990. [13] D. Wen, Y. Yuan, and J. Ou, “Monitoring the three-dimensional ionospheric electron density distribution using GNSS data over China,” J. Earth Syst. Sci., vol. 116, no. 3, pp. 235–244, Jun. 2007. [14] D. Wen, Y. Yuan, and J. Ou, “Three-dimensional ionospheric tomography by an improved algebraic reconstruction technique,” GPS Solutions, vol. 11, no. 4, pp. 253–258, Nov. 2007. [15] Y. Yuan, D. Wen, J. Ou, X. Huo, R. Yang, K. Zhang, and R. Grenfel, “Preliminary research on imaging the ionosphere using CIT and China permanent GNSS tracking station data,” in Proc. Int. Assoc. Geodesy Symposia, P. Tregoning and C. Rizos, Eds., Cairns, Australia, Aug. 22–25, 2005, vol. 130, pp. 876–883. [16] Y. Yuan and J. Ou, “The effects of instrumental bias in GNSS observations on determining ionospheric delays and the methods of its calibration,” Acta Geodaetica et Cartographica Sinica, vol. 28, pp. 110–114, 1999. [17] Y. Yuan, X. Huo, and J. Ou, “Models and methods for precise determination of ionospheric delays using GNSS,” Prog. Nat. Sci., vol. 17, no. 2, pp. 187–196, 2007. [18] D. Wen, Y. Yuan, J. Ou, K. Zhang, and K. Liu, “A hybrid reconstruction algorithm for 3-D ionospheric tomography,” IEEE Trans. Geosci. Remote Sens., vol. 46, no. 6, pp. 1733–1739, Jun. 2008.
Debao Wen received the B.Sc. degree in surveying engineering from Henan Polytechnic University, Jiaozuo, China, and the Ph.D. degree in geodesy and surveying engineering from the Chinese Academy of Sciences (CAS), Wuhan, China, in 2008. He is currently with the Changsha University of Science and Technology, Changsha, China, and with the Institute of Geodesy and Geophysics, CAS, Wuhan, China. His current research projects include 3-D ionospheric/atmospheric tomography based on GNSS observations and the effects of space weather on navigation and positioning. Dr. Wen was the recipient of several awards, including the Second Award for National Science and Technology Progress of China, the Second Award of the Chinese Military Science and Technology Advance, the Excellent Prize of the Presidential Scholarship of CAS, and the Student Best Paper Award of ION GNSS 2007.
3217
Yunbin Yuan received the B.Sc. degree in surveying engineering from the Shandong University of Science and Technology, Qingdao, China, in 1995 and the Ph.D. degree in geodesy and surveying from the Institute of Geodesy and Geophysics, Chinese Academy of Sciences (CAS), Beijing, China, in 2002. He is currently with Institute of Geodesy and Geophysics, CAS. His main subjects include satellite geodesy and GNSS positioning/atmosphere/ ionosphere. His current research interests are GNSSbased spatial environmental monitoring and analysis, including the ionospheric correction for Network RTK for centimeter-level realtime positioning; high-precision GNSS satellite navigation and positioning in different augmentation systems such as WAAS and Network RTK; GNSS in applications to orbit determination for Low Earth orbiter satellites; and China’s manned space engineering. Dr. Yuan was the recipient of several awards, including the Prize of Chinese Military Science and Technology Advance, Prize of Significant Contributions to China’s Manned Spaceflight, China’s Best Ph.D. thesis, Special Prize of the President Scholarship (CAS), and the 2001 ION GPS Student Best Paper Award.
Jikun Ou received the B.Sc. degree in mechanical engineering from the Beijing Machine College of China, Beijing, China, in 1970 and the M.Sc. and Ph.D. degrees in geodesy from the Institute of Geodesy and Geophysics, (IGG) Chinese Academy of Sciences (CAS), Beijing, in 1982 and 1994, respectively. He is currently a Research Professor with IGG, and a Doctoral Supervisor of the Graduate University, CAS. He is an Editor of five journals. Since 1980, he has been performing research on the theory of errors in geodesy and surveying. He is the author of more than 100 scientific and technical research articles. He has obtained a special allowance from the State Council of China since 1993, and was chosen as a specialist of outstanding contributions in Hubei province. His current major areas of expertise include the control of data quality, detection of gross errors, precise positioning by GPS, atmosphere and its effects on GPS surveying, and the approach of precise orbit determination for low orbital satellites and GNSS satellites, and so on.
Kefei Zhang received the B.Sc. and M.Sc. degrees in geodesy from Wuhan University (formerly Wuhan Technical University of Surveying and Mapping), Wuhan, China, in 1985 and 1988, respectively, and the Ph.D. degree in geodesy from the Curtin University of Technology, Perth, Australia, in 1997. Prior to joining the Royal Melbourne Institute of Technology (RMIT) University, Melbourne, Australia, in 1999, he was with the Institute of Engineering Surveying and Space Geodesy, University of Nottingham, Nottingham, U.K., as a Postdoctoral Researcher for about three years. He is currently an Associate Professor and leads the Measurement Science Group with the School of Mathematical and Geospatial Sciences, RMIT. He has a mixed background of satellite positioning, tracking, and geodesy. He is also a Guest Professor with a few international universities. His current research projects primarily involve algorithm development and innovative applications of GPS and GNSS technologies for high-accuracy positioning, atmospheric and meteorological studies, network-based RTK, and people and object tracking. In addition, he is also interested in geodesy-related theoretical and practical studies such as Earth’s gravity field modeling, satelliteorbit determination, and spectral and fractal analyses. He has been the author of about 100 peer-reviewed publications in these fields since 1990. Dr. Zhang is a member of a number of international special study groups and editorial boards of scientific journals. He is the President of the International Association of Chinese Professionals in Global Positioning Systems.
Authorized licensed use limited to: CHANGSHA UNIV OF SCIENCE AND TECHNOLOGY. Downloaded on July 19,2010 at 03:23:30 UTC from IEEE Xplore. Restrictions apply.
