Ionospheric Disturbances Triggered by China's Long March 2D Rocket

Using the data of 382 ground global navigation satellite system (GNSS) network stations in Western China, we studied and analyzed the ionospheric disturbances triggered by the “Long March” 2D rocket launch in Jiuquan, China on December 3, 2017. As compared with previous research, a higher sampling resolution for GNSS data (with a frequency of 1 Hz) was used to obtain more accurate occurrence times and propagation velocities of ionospheric disturbances. By using a method based on a quadratic function of time to fit a raw total electron content (TEC) series, a filtered TEC series was calculated using carrier observations, and a two-dimensional disturbances map was drawn. A new method, which accounts for the flight time of the rocket, was used to calculate the velocity of the shock wave. Ionospheric depletions and the shock wave were observed after the launch of the rocket. The depletion was observed within 100 to 1000 km south of the launch site along the rocket trajectory, which had a maximum amplitude of ∼3.8 TEC units (TECU), reaching ∼56% of the background TEC. A shock wave of V-shaped disturbances with amplitudes of ∼0.67 TECU was detected on both sides of the rocket trajectory. The shock wave moved southeast at an average velocity of ∼1861 m/s at a location 2200 km away from the launch site. Ionospheric disturbances with distances of more than ∼3000 km from the launch site were also observed.

Traveling ionospheric disturbances (TIDs) have been reported following numerous rocket launches, and reveal their characteristics within approximately 2000 km from the launch site. Calais and Minster [6] detected the shock wave signature of ionospheric disturbances following a space shuttle ascent using a single ground-based global positioning system (GPS) station. Afraimovich et al. [14] used a global international GNSS service (IGS) network to determine that the ionospheric response has the character of an N-wave, corresponding to the form of a shock wave as excited from launches of proton and space shuttle launch vehicles (LVs). A clear shock wave signature was also observed from a North Korea rocket launch, using the dense GPS networks in Japan [11]. Ding et al. [10] also found a V-shape shock acoustic wave excited by the launch of the long march 2F rocket using the dense GPS networks over China. These studies show that the TIDs generated by rocket launches are characterized by a V-shape shock wave signature near the trajectory, with horizontal velocities of 600-2500 m/s. In addition to the shock and acoustic waves, Lin et al. [15] and Liu et al. [12] found concentric TIDs (CTIDs) triggered by the launch of a SpaceX Falcon 9 rocket and China's Long March 4B rocket.
In addition to TIDs and CTIDs, ionospheric electron density depletion is another ionospheric disturbance excited by a rocket's blast plume, via interactions between perturbed neutral particles and ionized components [16]. Ozeki and Heki [9] found ionospheric holes caused by ballistic missiles from North Korea, as detected with a dense Japanese GPS array. Choi and Kil [17] used GNSS stations in South Korea to detect the large ionospheric total electron content (TEC) depletion induced by a 2016 North Korean rocket. Liu et al. [12] observed an ionospheric depletion excited by two launches of the Long March 4B rocket that carried "China-Brazil Earth Resources Satellite 3" and "China-Brazil Earth Resources Satellite 4," respectively. These studies show that the TEC series shows an evident decrease several minutes after the launch of the rocket. The TEC depletion can reach ∼50% of the background TEC, and can last 1-2 h.
In this article, the ionospheric disturbances triggered by the launch of the long march two-dimensional (LM-2D) rocket are analyzed for the first time, using the dense ground GNSS network in China. From this launch event, we observed two different types of ionospheric disturbances: ionospheric depletions and shock waves. The ionospheric depletion main appeared at 300 to 400 km away from the launch site, and the maximum amplitude of the depletion can reduce 56% of the background value of the TEC. The ionospheric shock waves clearly appeared after the rocket launch along the rocket trajectory, even 3000 km far away from the launch site, and its average velocity can reach ∼1861 m/s.

A. Event Information
The LM-2D is a two-stage carrier rocket, and is mainly used for launching low-earth orbit (LEO) and sun synchronous orbit (SSO) satellites. At 04:11 universal time (UT) [local time (LT) = UT + 8] on December 3, 2017, an LM-2D LV successfully launched the "Land Survey Satellite-1" (LKW-1) at the Jiuquan Satellite Launch Center (40.96°N, 100.29°E, red star in Fig. 1). Fig. 1 also shows the projection (blue line) of the rocket's approximate flight trajectory on the ground, according to the Federal Aviation Administration (https://pilotweb.nas.faa.gov/ Pilot-Web/). After launching, the rocket flew southwest with an azimuth (clockwise from North) of the trajectory at approximately 194°, passing through Myanmar, the Bay of Bengal, and the Andaman and Nicobar Islands, and finally entering the Indian Ocean. The rocket traveled southwest for more than 3000 km, reaching an apogee of 500 km altitude, achieved a velocity of more than 7 km/s, and sent the spacecraft to a LEO of 488 by 504 kilometers, inclined at 97.46°. The LV completed its flight sequences, including first/second stage separation, fairing jettison, 2nd stage main engine shut down, 2nd stage vernier engine shut down, and spacecraft / LV separation at t0 + 154 s, t0 + 184 s, t0 + 328 s, t0 + 620 s, and t0 + 637 s, respectively, after the LV lifted off at 04:11 UT (t0).

B. Data Processing
The GNSS data used in this article are mainly obtained from the Crustal Movement Observation Network in China, the Satellite Navigation and Positioning Reference Station System of Yunnan Province and Qinghai Province in China, and the GNSS networks provided by the IGS. After analysis and data preprocessing, the datum of 382 stations is selected for analysis in this article. The geographical locations of the GNSS stations are shown in Fig. 1. To obtain more accurate characteristics of ionospheric disturbances, we adopted a higher time resolution data (1 s) from the GNSS stations.
First, the carrier phase observations and pseudo-range observations of the GNSS stations are subjected to preprocessing, such as cycle slip detection and repair, to obtain "clean" data. Then, using the GNSS dual-frequency observation data, a carrier phase smoothing method is used to calculate the slant TEC (sTEC). Based on a thin-layer ionosphere assumption, a modified single-layer mapping function is used to convert the sTEC into a vertical TEC (vTEC). The thin-layer ionosphere altitude and the elevation cutoff are set to typical values of 300 km [9] and 30°, respectively. To analyze the characteristics of the ionospheric depletion, we used a second-order polynomial method to fit the original vTEC series using (1), and estimated the values for a, b, and c using a least-squares method with 30 min of period data, before the rapid decrease of vTEC series. Then, we used (1) to calculate the value of the fitting TEC series, and the vTEC anomaly was the fitting TEC series minus the raw TEC series vTEC(t) = at 2 + bt + c.
(1) Fig. 2(a) shows the raw and ployfit TEC series of the station "qhmd," as received by satellite G20. The blue solid line and the red solid line represent the raw vTEC series and the ployfit vTEC series, respectively. An unusual TEC decrease with an amplitude of ∼3 TEC units (TECU) appeared 5 min after the launch of the rocket. After nearly two hours, the TEC value gradually returned to normal. Usually, dual-frequency GPS data with a time resolution of 30 s is used to study ionospheric disturbances, which has an error of ± 30 s in abnormal times, and larger noise in a vTEC series. Here, only the carrier phase observations with 1 s resolution are used to calculate the disturbance TEC (dTEC), to obtain more accurate characteristics of the ionospheric disturbances using (3) Here, sTEC is the slant TEC, L1 and L2 are the GNSS phase measurements, P1 and P2 are the GNSS code measurements, f1 and f2 are the frequencies of the GNSS (f1 = 1575.42 MHz and f2 = 1227.60 MHz), ε P is the sum of the hardware delay and observed noise for a pseudorange observation, and ε L is the sum of the ambiguity of the carrier phase observation, hardware delay, and observed noise ε P and ε L are regarded as constants during the calculation. We applied a zero-phase sixth-order finite impulse response Butterworth filter, with cutoffs at 2 and 12 min, to determine the sTEC L . The constant terms ε P and ε L can be estimated by (2) and (3). Then, we converted the fitting sTEC to a vTEC. As shown in Fig. 2(b), the filter vTEC P and the filter vTEC L are approximately equal in size and opposite in sign, and have X-axis symmetry. Fig. 2(b) shows that the times of the maximum amplitude of the filter vTEC P and filter vTEC L values are at 330 and 324 s after launch, respectively. The maximum amplitudes are 0.53 TECU and −0.51 TECU, respectively. Therefore, the inverse sign of the vTEC L series can be used to analyze and obtain more accurate characteristics of the ionospheric disturbances.  launch, and the amplitudes reach −0.05, 0.3, 0.66, 0.67, 0.53, 0.33, and 0.52 TECU, respectively. When these TEC series begin to have perturbations, they exhibit an N-shaped signature (except for the station gsgt), with a maximum amplitude of ∼0.67 TECU. The first and last stations with positive perturbations are station zysn and pbri, respectively, and the IPPs are located 140 km and 3000 km south of the launch site, respectively. Although the station pbri is the farthest from the launch site, its perturbation amplitude still reaches ∼0.52 TECU. Fig. 4(b) shows the TEC perturbation series observed from seven stations (see Fig. 1, blue triangles) 480 km away from the launch site from the west to the east, i.e., perpendicular to the trajectory. The first wave peaks of the TEC perturbation series are reached at 515, 475, 387, 270, 311, 432, and 463 s after the launch, and the amplitudes reach 0.07, 0.13, 0.39, 0.61, 0.48, 0.13, and 0.08 TECU, respectively. The maximum amplitude of the TEC perturbations of other stations decreases gradually from the center of the station bfgh to the east-west side, and the time of its maximum amplitude gradually increases from the east-west side. The periodic oscillation perturbations after the N-shaped perturbation at the station xunh and lxjs, on the east side of the rocket trajectory, are more evident than those at other stations. Fig. 4(c) shows the dTEC series for seven stations (see Fig. 1, red triangles) 1800 km away from the launch site from the west to the east, i.e., perpendicular to the trajectory. The first wave peaks of the TEC perturbation series are reached at 629, 696, 724, 773, 807, 850, and 864 s after the launch, and the amplitudes reach 0.2, 0.2, 0.15, 0.08, 0.05, 0.04, and 0.03 TECU, respectively. The locations of the IPPs with the maximum amplitude of the station tobg are closest to the trajectory, and the amplitudes of other stations gradually decrease from west to east. The time of the maximum amplitude appears to gradually increase from west to east.

A. Shock Wave Propagation
As the IPPs are farther from the trajectory, the amplitude of the perturbation is gradually reduced to the noise level of the TEC signal (∼0.03 TECU). The aforementioned results indicated that the closer the location of the IPPs to the trajectory, the larger the amplitude of the first wave peak of the perturbation, and the shorter the time after the launch of the rocket. The perturbation presents as N-shaped signature (the maximum amplitude can reach ∼0.67 TECU), followed by a periodic oscillation (the amplitude ranges from 0.05 to 0.1 TECU). This indicates that the ionospheric disturbances are generated near the rocket trajectory, and then gradually spread to the east and west sides of the rocket trajectory before disappearing.
2-D maps of TEC perturbations have extensive used to analyze the characteristics of the ionospheric disturbances [18], [19], [20]. By analyzing the propagation trend of the ionospheric disturbances as shown in Fig. 5, the shock wave first appeared 210 s after rocket launch near the launch site, and an obvious V-shaped shock wave gradually appeared 390s after rocket launch along the rocket trajectory at 300 to 700 km south of the rocket launch site. The shock wave with evident positive and negative phase structure characteristics gradually propagated southeast at 1800km away from the launch site. In the process of its propagation to a perpendicular distance of 600 km from the trajectory, the amplitude of the disturbance gradually decays until it disappears, and its duration can reach 990 s+ after the launch of the rocket [see plot (28) in Fig. 5].
To further analyze the propagation characteristics of the ionospheric disturbances, a time distance map is plotted as a function of the perpendicular distance of the IPPs to the rocket trajectory, and the time of the IPPs for two regions. The results of Fig. 6(a) and (b) were calculated using the IPPs of region A and region B (see Fig. 1), respectively. The method for calculating the propagation velocity uses the distance perpendicular to the trajectory [10]. Fig. 6(a) shows that the shock wave can be observed on both sides of the rocket trajectory. The maximum distance of the ionospheric disturbances to the west and east is ∼300 km. The propagation velocity of the shock wave is 835 and 973 m/s to the west and east, respectively. The velocity of the waves gradually decreases as the waves move away from the rocket. Fig. 6(b) shows that the shock wave can only be observed on the east side of the rocket track, as there are no IPPs on west of the trajectory. The maximum distance of the eastward propagation of the ionospheric disturbances is ∼750 km, and the propagation velocity of the waves is 2211 m/s (without considering the rocket velocity). Fig. 6(a) shows that the duration of the ionospheric disturbances is longer than the result shown in Fig. 6(b), which is consistent with the results shown in Figs. 4 and 5.
To obtain the excitation source and more accurate propagation velocity of the shock wave, high sampling data of 1 s is used to analyze. We selected the locations and occurrence times of the largest perturbations (IPPs) of the same satellite, close to the rocket track in region B (see Fig. 1). Here, the ionospheric disturbances are more evident and the IPPs are more dense, which is useful for calculating and analyzing. The propagation velocity of the shock wave is calculated according to the way the excitation source directly propagates to the IPPs [6]. Fig. 7 shows a schematic diagram of the rocket's flight, as well as the shock wave propagating from the excitation source to the IPPs. It clearly shows the process of the shock wave generated during the rocket launch.
Based on the assumption that the shock wave was excited by the rocket's supersonic speed movement [10] and propagated from the source to the IPPs [6], we design three calculation schemes for analyzing the excitation sources of the ionospheric disturbances, and calculate the propagation velocity of the shock wave.
In scheme 1, it is considered that there is only one excitation source, located at a certain point on the rocket trajectory. As is shown in Fig. 8(a), R 0 is the excitation source of all IPPs (P i ), the time of the first largest perturbation of point P 0 is t 0, the time of the first largest perturbation of point P i is t i, and S i is the distance from the excitation source to the IPPs (P i ). The propagation speed V i of the shock wave can be calculated according to (4). Starting from the perpendicular foot F 0 , the excitation source searches along the trajectory of the rocket toward the launch site in 10 km steps. The average velocity with the lowest standard  deviation of each velocity is taken as the propagation velocity of the shock wave In scheme 2, there are multiple excitation sources, and the excitation source is the perpendicular foot of the IPPs in the rocket trajectory [see Fig. 8(b)]. As the excitation source is assumed to be the perpendicular foot of the IPPs in the rocket trajectory, the speed of the shock wave can be calculated directly according to (4), without searching for the excitation source.
In scheme 3, a special case is considered, i.e., the rocket flight time in scheme 2. The flight distance D is approximately equal to the projection of the distance between the IPPs perpendicular to the rocket trajectory. Assuming that the velocity of the rocket is V r , the speed of the shock wave can be calculated directly, according to Fig. 9(a) shows the standard deviation of the shock wave propagation velocity, as calculated by scheme 1, scheme 2, and scheme 3 for region B, respectively. The black solid line, blue solid line, and red dotted line represent the standard deviation results for the velocity calculated in scheme 1, scheme 2, and scheme 3, respectively. The altitude of the IPPs is calculated according to an interval of 100-600 km and a step length of 20 km. Fig. 9(a) shows that the minimum standard deviation of the shock wave propagation speed in schemes 1 and 2 is over 400 m/s in the range of 100-600 km, whereas the difference in the shock wave propagation speed as calculated in scheme 3 is not significant at different IPPs altitudes, and the standard deviation of the speed at 460 km reaches the minimum (±80.1 m/s). In (5), the velocity V r of the rocket is unknown. Taking a rocket velocity of 5-8 km/s as the range, and searching in steps of 0.1 km/s, the propagation velocity V i can be calculated. Fig. 9(b) shows that the standard deviation of the shock wave propagation velocity, as calculated using different rocket speeds, is the smallest when the rocket speed is 6.3 km/s, and the average speed of the shock wave is 1861 m/s when the IPPs altitude set to 460 km. Fig. 9 shows that the bias of the propagation velocity of the shock wave in region B is minimum in scheme 3. Therefore, the results are supported to the shock wave may be the multi-point excitation from the rocket motion and the calculation of the propagation velocity must consider the rocket velocity. Fig. 10(a)-(c) shows the TEC series of satellites G20, G21, and G29 observed from seven stations, respectively. Fig. 10(a) illustrates the TEC series observed from stations along the rocket trajectory, whereas Fig. 10(b) and (c) illustrate the TEC series observed from stations 480 and 1800 km away from the launch site and perpendicular to the rocket trajectory, respectively. The red dotted line denotes the rocket launch time, the blue solid line denotes the raw vTEC series, and the red solid line denotes the fitting vTEC series, as calculated by (1). Fig. 10(a) shows that the depletion appeared evidently from the station gjtj, and reached a maximum of ∼3.6 TECU at the station qhmd, before gradually decreasing to ∼1.1 TECU at the station pbri 3000 km away from the launch site. Fig. 10(d) shows the distributions and sizes of the depletions of the seven stations along the rocket launch trajectory. It clearly reflects that the depletion changes from small to large, then from large to small. Fig. 10(b) shows that the amplitude of the depletion reaches the maximum value near the trajectory, and decreases farther from the trajectory. Fig. 10(c) shows that there is almost no depletion in the TEC series of these stations. Perpendicular to the trajectory, Fig. 10(c) also clearly shows that the depletion near the trajectory reaches the maximum, and that the depletion gradually spreads to both sides and decreases. Fig. 11 shows a 2-D snapshot of the space depletion distribution (with intervals of 1, 10, and 20 min) from 5 to 140 min after the launch, clearly indicating the dynamic change process of the dTEC with time and space. Five min after the launch, the TEC depletion first appears, 300 km south of the launch site [see Fig. 11(a)]. The depletion area gradually expands mainly around the rocket launch track, 200-800 km away from the launch site [see Fig. 11(a)-(h)]. From 30 to 40 min after the launch, the range of the depletion distribution reaches a maximum, covering between 100 and 1000 km away from the launch site along the trajectory, and the width is ∼300 km perpendicular to the trajectory (and is almost symmetrical to the trajectory). The depletion is mainly concentrated in the range of 400 to 600 km south of the rocket launch site. The amplitude of depletion reaches the maximum at ∼3.8 TECU. For 40 min after the launch of the rocket, the depletion range gradually narrowed down until it subsided, and the entire process lasted ∼2 h. In that regard, 120 min after launch, the depletion could hardly be observed.

A. Shock Wave
According to the rocket flight video and GNSS observation data of the rocket launch site, Fig. 12(a) draws the relationship diagram of the rocket's flight time, flight height (blue curve), and flight distance (red curve) after launch. Fig. 12(b) shows the relationship between the flight distance, flight altitude (blue curve), and flight speed (red curve) after launch. The black asterisk, the black triangle, and the circle represent three important moments, respectively: the first and second stage separation (t0 + 154 s), the second stage shutdown (t0 + 328 s), and the separation of the rocket and satellite (t0 + 637 s). The LV of this launch event is an SSO satellite, with a flight altitude of ∼500 km. The rocket passes through the D-F2 layer of the ionosphere. Fig. 12 shows that when the rocket is 500 and 1500 km away from the launch site, the flight times are t0 + 300 s and t0 + 460 s after launch, respectively, and the altitudes reach 200 and 400 km, respectively.
According to Fig. 12, the second stage main engine of the rocket has been shut down (t0 + 328 s), the speed has reached 6 km/s, and the distance from the launch site has exceeded 1500 km. The rocket is still rising, and releases most of the thrust when it is in the initial stages of launch (t0 + 200 s -t0 + 300 s). The rocket flies nearly horizontal at supersonic speed when it is in the middle and late stages of launch (t0 + 400 s -t0 + 500 s). The excitation source is located in the rocket trajectory, and moves dynamically with the rocket. The engine tail flame of the rocket can no longer produce large acoustic explosion disturbances.
The velocity of the shock wave is 800-1000 m/s and 2210 m/s in region A and region B are shown in Fig. 6, respectively. In scheme 3, considering the rocket velocity, the horizontal velocity of the shock wave is 1861 m/s in region B. The velocity of the shock wave in region A is coincident with the other research results in the velocity range of 600-1670 m/s [10], [11], [12], [21]. However, the velocity of the shock wave in region B is higher than the previous research results. Only Kakinami et al. [22] who used a dual-frequency GNSS to observe a missile launch in North Korea, calculated a result that exceeded 2000 m/s. Based on the analysis of previous launch events, it is found that the flight altitude of most of the previous research events is only ∼200 km [4], [5], [6], [8], [10] and that most of the excitation sources of the shock wave are at altitudes of +100 km. In one SSO satellite launch event [12], although the altitude of the rocket reached ∼700 km, the analyzed and calculated IPPs were only ∼1500 km away from the launch site that the rocket flight altitude had still not reached 400 km, i.e., the excitation source of the shock wave had not reached 400 km. Therefore, the propagation speed of the shock wave is slightly higher than that shown in previous studies, which may be owing to the fact that the flight altitude of the rocket in area B exceeds 400 km, and because the propagation speed of a shock wave generated by rocket at a higher atmospheric altitude is higher.
In addition, by calculating the distance from the IPPs to the satellite launch site, the velocity of the rocket can be roughly calculated. Fig. 13 shows the time distance map drawn by the series of satellite G20 observed from the stations of the pink triangle (see Fig. 1), and indirectly reflecting the rocket speed. The mean speeds of the rocket are 1.2, 3.8, and 6.8 km/s at 300 km, 1200 km, and 3000 km from the launch site, respectively, reflecting the gradual acceleration of the rocket. The calculation results are consistent with those in Fig. 12. The depletion first appears 5 min after the launch, 300 km away from the launch site. At this time, the rocket has achieved first/second stage separation, and the altitude of the rocket is ∼100 km. After that, the depletion gradually increases, and the central area of the depletion is 400 km south of the rocket launch site, reaching the maximum value. With the secondary main engine shut down (t0 + 328 s), the distance from the south of the launch site to 700 km along the trajectory is gradually reduced, and the depletion gradually decreases, disappearing after the distance from the south of the launch site reaches 1000 km. Before the first/second stage separation, the depletion is not evident that the altitude of the rocket does not reach 100 km. After the second stage shutdown of the rocket, the matter ejected into the atmosphere by the rocket decreases, and the depletion gradually decreases.

B. Depletions
It is concluded that the formation time of depletion is mainly between the time nodes of the first/second stage separation and the second stage shutdown of the rocket. At this time, the altitude of the rocket reaches 100-300 km, and the rocket is in an accelerating/rising stage, which consumes a significant amount of chemical fuel, and releases a significant amount of combustion propellant into the ionosphere. The combustion of the rocket tail flame reacts with the chemical reactions of other ions in the atmosphere (a large number of H 2 O and H 2 ions, and a large number of O + ions occupying the F2 layer). Then, these molecular ions combine and react rapidly again in the ionosphere, forming a hole in the ionosphere. The depletion has an amplitude of ∼3.8 TECU, which is smaller than the ∼12 TECU found in other studies [12]. The reason is that the background TEC is only 6.8 TECU, and the reduction ratio of depletion is still 56%, consistent with the previous research results of 50% [23]. In addition to the influence of the amount of combustion gases emitted by different launch missions, the depletion is also related to the TEC value of the background field.
The depletion can last for ∼2 h, longer than found in the previous observation results (i.e., 1 h) [9], [24], [25], but more consistent with the observation results of Liu et al. [12]. The spacecraft orbit altitude of the former is generally less than 200 km, but the latter is similar to this launch mission (SSO), with an altitude of more than 500 km. The main reason is the difference in the amount of propellant consumed by different LVs via injection into the atmosphere, and the altitude of the LV. The key time nodes of the rocket launch are consistent with the amount of propellant consumed by injection into the atmosphere.
In this launch event, the depletion area is roughly symmetrical to the rocket trajectory, which is inconsistent with the observation records, with evident westward deviation [12], [17]. According to the horizonal wind model HWM14, the neutral wind speed in the area with the largest depletion area (400-600 km south of the launch site) is ∼3.6 m/s. Owing to the smaller speed of the background wind, the depletion basically diffuses evenly around the rocket track, without a large deviation. Choi and Kil [17] and Liu et al. [12] used HWM07 and HWM14 models to calculate the wind speed in the depletion area, reaching 120 and 70-90 m/s, respectively. The background wind has a large westward component, so the depletion presents a significant west bias phenomenon. The observation results and wind speed results also support the view that the movement of the depletion may be related to the background winds.

C. Conclusion
In the article, we investigated the effects of the launch of the rocket on the ionosphere. The ionospheric depletions were obtained by a second-order polynomial fitting of the vTEC observations. The depletion main appeared at 300-400 km south of the launch site as the center, and gradually symmetrical spreads around along the rocket trajectory, lasting for ∼2 h. The maximum amplitude of the depletion was ∼3.8 TECU, accounting for 56% of the background value of the TEC. In the process of rocket launching, the large amount of water vapor produced by fuel consumption reacts with other ions in the atmosphere, which is the main cause of ionospheric holes. In addition, the shock wave is obtained by a band-pass filter of the vTEC L with a high sampling rate of 1 s observations. The shock wave first appeared at 140 km away from the launch site after the launch, and still be observed 3000 km away from the launch site. The average velocity of the shock wave can reach ∼1861 m/s when the shock wave appeared 2200 km away from the launch site at an altitude of more than 400 km. The propagation speed of a shock wave at a higher altitude is higher than at a lower altitude. The excitation source of the shock wave may be the multi-point excitation from the rocket motion, and the supersonic motion of the rocket is the main cause of the shock wave.