Latitudinal variations of middle thermosphere: Observations and ...https://www.researchgate.net/.../Latitudinal-variations-of-middle-thermosphere-Observati...

JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 116, A12306, doi:10.1029/2011JA017067, 2011

Latitudinal variations of middle thermosphere: Observations and modeling Jiuhou Lei,1 Jeffrey M. Forbes,2 Han‐Li Liu,3 Xiankang Dou,1 Xianghui Xue,1 Tao Li,1 and Xiaoli Luan1 Received 13 August 2011; revised 2 September 2011; accepted 16 September 2011; published 6 December 2011.

[1] Ching and Carter (1974) reported an equatorial bugle of midnight mass density at about 230 km on the basis of ion gauge measurements in April 1972. However, since then few succeeding studies of the variation of equatorial mass density in this region have been pursued. In this report, observations from the Satellite Electrostatic Triaxial Accelerometer (SETA) satellite during 1 September 1983 to 31 October 1983 are used to examine the variations of the equatorial density at around 210 km for the daytime (1030 LT) and nighttime (2230 LT) sectors. It is found that mass densities show an equatorial trough around the geographic equator and two crests aside the trough during both daytime and nighttime. The thermosphere‐ionosphere‐mesosphere‐electrodynamics general circulation model (TIMEGCM) reproduces the features seen in the SETA data. Our simulation reveals that the observed double‐hump structures in both daytime and nighttime sectors are mainly associated with the tides excited locally in the lower and middle thermosphere region, although they are modulated by the tides originating from the lower atmosphere. Citation: Lei, J., J. M. Forbes, H.‐L. Liu, X. Dou, X. Xue, T. Li, and X. Luan (2011), Latitudinal variations of middle thermosphere: Observations and modeling, J. Geophys. Res., 116, A12306, doi:10.1029/2011JA017067.

1. Introduction [2] Ching and Carter [1974] reported a magnetic equatorial bugle of mass density at around 230 km on the basis of ion gauge measurements in April 1972. Their observations suggested geomagnetic control of the equatorial maximum in mass density, which is in accordance with the N2 distribution from OGO‐6 data [Philbrick and McIsaac, 1972; Hedin and Mayr, 1973]. Further analysis is obviously required to understand the latitudinal structure of equatorial mass density in the middle thermosphere region given that Ching and Carter draw their conclusion from a limited data set, i.e., 13 orbits during a 2 day period. As noted by Ching and Carter [1974], an analysis of other data is necessary to improve our knowledge and understanding of the variation in equatorial mass density. [3] Nevertheless, unlike ionospheric data, historic thermosphere density measurements are very sparse. To our knowledge, there have not been many studies of the variation of equatorial mass density in the middle thermosphere since 1974, although AE‐E [Herrero and Spencer, 1982]

and CHAMP [Liu et al., 2005; Ma et al., 2010] data were used to examine the variations of nighttime mass densities in the upper thermosphere. Exceptions include Arduini et al. [1997] and Akmaev et al. [2009, 2010], whereas these studies did not focus on the features we will discuss afterward. Arduini et al. [1997] presented an extensive analysis of the equatorial nighttime density variability based on data from San Marco III and V satellites flying in low‐inclination orbits in 1971 and 1985, respectively. Their analysis also extends down to 220 and 260 km in the middle thermosphere. Akmaev et al. [2009, 2010] showed the Whole Atmosphere Model (WAM) simulations of low‐ to middle‐ latitude density variations down to the lower thermosphere. In this brief report, observations from the Satellite Electrostatic Triaxial Accelerometer 3 (SETA3) satellite [see Forbes et al., 1995; Rhoden et al., 2000] in 1983 and numerical simulation from the thermosphere‐ionosphere‐ mesosphere‐electrodynamics general circulation model (TIMEGCM) [Roble and Ridley, 1994] are used to study the variations of middle thermosphere density.

2. Observations 1

School of Earth and Space Sciences, University of Science and Technology of China, Hefei, China. 2 Department of Aerospace Engineering Sciences, University of Colorado at Boulder, Boulder, Colorado, USA. 3 High Altitude Observatory, National Center for Atmospheric Research, Boulder, Colorado, USA. Copyright 2011 by the American Geophysical Union. 0148‐0227/11/2011JA017067

[4] Mass densities in the altitude range between 170 and 240 km obtained from the SETA3 satellite are used in this study. SETA3 operated in a near circular, sun synchronous orbit from July 1983 to March 1984, providing both daytime (1030 LT) and nighttime (2230 LT) data at low‐middle latitudes. Observations near equinox conditions in 1 September 1983 to 31 October 1983, including 38 days of data, are utilized in the current study. The mean solar activity F107

A12306

1 of 7

A12306

LEI ET AL.: LATITUDINAL VARIATIONS OF THERMOSPHERE

A12306

magnetic control of the equatorial density was found on other days during our study period. [6] The zonally averaged mass densities at 210 km during 1 September 1983 to 31 October 1983 are depicted in Figure 3, in which each curve represents the relative density on an individual day with respect to the density at the geographic equator. In this figure, the measured densities r(z) at satellite altitudes are normalized to the mean satellite altitude at 210 km using both NRLMSISE00 and JB2006 [Bowman et al., 2008] and then scaled by the densities at the equator. These two empirical models are used in this normalization process in order to assess whether the hump structure is caused by the transformation of data to the reference altitude, given that the NRLMSISE00 model predicts similar crests in mass density but the JB2006 does not (see Figure S1).1 As seen in Figure 3, the densities normalized by the JB2006 are similar to those by the NRLMSISE00, which indicates that the two hump structure is not associated with the altitude normalization process. At 1030 LT (Figures 3a and 3c), there are two crests occurring at around ±20°. On some days, a secondary crest can be seen at around 60–70° in the Northern Hemisphere. At 2230 LT (Figures 3b and 3d), the northern crest locates within the latitudinal range of 20–40°N, whereas the

Figure 1. Mass densities normalized at 210 km at (a) 1030 and (b) 2230 LT observed by the SETA3 on 30 September 1983. The red dashed lines represent the mean density in each latitude bin. was about 110. The measured densities r(z) at satellite altitudes are normalized to the mean satellite altitude at 210 km using NRLMSISE00 [Picone et al., 2002], as follows: r(210) = r(z) × rM(210)/rM(z), where rM(210) and rM(z) stand for the mass densities calculated from the NRLMSISE00 model at 210 km and satellite altitudes, respectively. [5] Figures 1a and 1b show the normalized mass densities on 30 September 1983 on the dayside and nightside, respectively, as an example. The red dashed lines represent the mean density as a function of geographic latitude. The predominant feature on this day is the two hump structure that is seen at low‐middle latitudes in both the morning and evening sectors. The crests are located at ±20° at 1030 LT, whereas they sit at ±40° at 2230 LT. The trough appears to be organized by the geographic equator. To further demonstrate the geographic control of the equatorial thermosphere density, the SETA densities along all orbits on 30 September are shown in Figure 2 as a function of longitude and latitude. It appears that the minimum density at the equatorial region generally corresponds better with the geographic equator than the geomagnetic equator. Note that no obvious geo-

Figure 2. Variation of mass density (in unit of 10−10 kg m−3) at 210 km at (a) 1030 and (b) 2230 LT along the SETA3 orbits on 30 September 1983. Note that the equatorial trough in mass density is not aligned with the dip equator (dashed line).

1 Auxiliary materials are available in the HTML. doi:10.1029/ 2011JA017067.

2 of 7

A12306


A12306

Figure 3. Ratios of the observed mass densities at 210 km with respect to the corresponding density at the geographic equator during 1 September 1983 to 31 October 1983. Both the NRLMSISE00 (left) and JB2006 (right) models are used to transfer the data to the reference altitude.

southern crest appears at 40–50°S. Additionally, the crest‐ to‐trough ratios are in the range between 1.05 and 1.15, with slightly higher ratio at 2230 LT.

3. Model Simulation [7] In order to understand the cause of the observed latitudinal structure of the equatorial density in Figures 1–3, we conducted numerical experiments with the NCAR‐TIMEGCM [Roble and Ridley, 1994]. In this study, the horizontal resolution of the TIMEGCM was 2.5° × 2.5°, and the vertical resolution was one quarter of the scale height. The numerical experiments were conducted in equinox and under the similar geophysical condition as the SETA3 data. Specifically, the solar activity index F107 is 110 and the geomagnetic activity condition is quiet. Akmaev et al. [2009] and Ma et al. [2010] demonstrated that the formation of the midnight temperature or density maximum in the upper thermosphere is associated with the tidal effect in the lower thermosphere, but they did not address whether the associated tides are excited in the MLT region or propagate from the troposphere. In this study, we conducted a TIMEGCM simulation first without tidal forcing at the model lower boundary at 30 km. This simulation can serve as a baseline, and the corresponding results are shown in Figure 4.

[8] Figure 4a shows the simulated mass density at 210 km as a function of latitude and local time. It is obvious that the latitudinal variations of density in the equatorial region vary with local time. The vertical dashed lines illustrate the two local times measured by the SETA3 satellite in 1983. The observed features in the SETA density are reproduced by the TIMEGCM. The simulated density at 2230 LT has a minimum at the geographic equator and two maxima at ±40°, whereas at 1030 LT two maxima of density locate at ±20°. Surprisingly, these results agree with our observations fairly well although the tidal forcing is not specified at the model lower boundary. [9] Figure 4b displays the ratios of simulated mass densities at 210 km to the density at the geographic equator for 1030 and 2230 LT, which provides a quantitative comparison between the TIMEGCM simulation and the SETA observations (Figure 3). The crest‐to‐trough ratios in Figure 4b are around 1.05 and 1.12 at 1030 and 2230 LT, respectively, which are in accordance with the observations in Figure 3. It should be pointed out that the TIMEGCM gives symmetric crests in two hemispheres in equinoctial season, but the observed densities are larger in the Northern Hemisphere at 1030 LT and in the Southern Hemisphere at 2230 LT. [10] Figure 4c shows the ratios of simulated mass densities at 2230 LT to the corresponding density at the geographic equator as a function of altitude and latitude. Again,

3 of 7

A12306


Figure 4. TIMEGCM simulation at 00 UT including (a) mass density at 210 km as a function of latitude and local time, (b) ratios of mass densities at 210 km to the density at the geographic equator for 1030 and 2230 LT, and (c) ratios of simulated mass densities at 2230 LT to the corresponding density at the geographic equator as a function of altitude and latitude. Note that in this model simulation no tidal forcing is applied at the lower boundary.

Figure 5. Same as Figure 4 but for the case with migrating tides specified at 30 km of the model boundary. 4 of 7

A12306

A12306


A12306

Figure 6. Longitudinal variations of ratios of mass densities at 210 km to the density at the geographic equator for (left) 1030 and (right) 2230 LT for different specification of tidal forcing at the model lower boundary: (a) without tidal forcing, (b) with migrating tides, and (c) with migrating and nonmigrating tides. Either migrating or nonmigrating tides include both diurnal and semidiurnal components in the simulations. All simulations show that the equatorial trough in the morning‐nighttime densities is not aligned with the dip equator (dash line).

the simulated densities show a symmetric structure in two hemispheres. The TIMEGCM density also varies with altitude, with lows around 70, 100, and 150 km, and highs around 80 and 120 km and above 190 km. A similar situation is also seen at 1030 LT, as shown in Figure 4d. This phase changing with altitude reflects well the tide effect on the thermosphere. As stated before, we did not include tidal forcing at the model boundary in this case. Therefore, the reflected tides in Figure 4 should be excited at the altitudes above 30 km. Nevertheless, the detail mechanisms and characteristics for these excited tides are beyond the scope of this study. [11] Next we carried out the TIMEGCM simulations by specifying the amplitudes and phases of tides from the lower

atmosphere from the Global Scale Wave Model (GSWM02) [Hagan and Forbes, 2002, 2003] at the lower boundary of the TIMEGCM. Both diurnal and semidiurnal tides with the modes up to wave number 6 are applied in the simulations. Figure 5 shows the results with the same format as Figure 4, but for the case with migrating tides only specified at 30 km of the model boundary. Clearly, the dominant features in Figure 5 are very similar to those seen in Figure 4, except for the differences in quantitative sense. This conclusion becomes more obvious in Figures 6a and 6b, in which longitudinal variations of ratios of mass densities at 210 km to the density at the geographic equator are compared between the simulations without tidal forcing and with migrating tides specified at the model lower boundary. Our

5 of 7

A12306


further sensitivity simulations suggested that nonmigrating tides have significant effects on the modulation of the longitudinal variations in mass density. As shown in Figure 6c, in this fixed local time frame there are four obvious peaks in the normalized densities with respect to the equatorial values, which might be associated with the upward propagation of the DE3 nonmigrating tides [Hagan et al., 2007]. The simulated results in Figure 6c can explain the longitudinal pattern see in the SETA observations in Figure 2. Thus, both migrating and nonmigrating tides propagating from the troposphere have significant modulation on the variations of mass density in the middle thermosphere, but they are not the major contributors for the formation of the two hump structure observed in the SETA density.

4. Discussion [12] As shown in Figures 1–3, the latitudinal structure in the equatorial density from the SETA3 differs with that from the ion gauge measurements reported in Ching and Carter [1974], although the data used in both studies were under equinox conditions and with similar solar activity level (F107 = 110). Ching and Carter [1974] showed an equatorial maximum in nighttime mass density instead of a trough, as seen in the SETA data. Note that Ching and Carter normalized their density at the constant altitude of 230 km, which is higher by 20 km than the reference altitude in this study. As revealed in the simulation result (Figure 4c), density shows a trough in the altitude range between 190 and 400 km. Therefore, the differences in altitude cannot explain the different features seen in these two data sets. [13] We should mention that there is the slight difference in the local time sampling of these two satellites. In the observations of Ching and Carter [1974] a difference of about 2 h in local time exists between 60°S and 60°N, and the nighttime local time at the equator was about 0130 LT, which is about 3 h later than that of the SETA3 satellite. As illustrated in Figure 4a, the TIMEGCM predicts an equatorial bugle in nighttime mass density during 0200– 0500 LT. However, the simulated bugle in the nighttime equatorial density is located at the geographic equator rather than at the geomagnetic equator as claimed in the study of Ching and Carter [1974]. Alternatively, the observed equatorial density bugle by Ching and Carter [1974] might be actually organized by the geographic rather than geomagnetic equator. [14] Finally, it is worth noting that the two hump structure has been found in the daytime equatorial thermosphere density and temperature, which is called the equatorial thermosphere anomaly (ETA), a similar phenomenon to the equatorial ionization anomaly (EIA). Besides the OGO6 data [Philbrick and McIsaac, 1972; Hedin and Mayr; 1973], DE2 neutral temperature [e.g., Raghavarao et al., 1991, 1993] and CHAMP thermospheric density [Liu et al., 2005, 2007; Lei et al., 2010] were used to study the characteristics of the ETA. All these studies revealed a strong geomagnetic control of the ETA in the upper thermosphere, which is suggestive of an ion neutral coupling connection. However, our results in Figure 6 demonstrate the geographic control of the equatorial trough in the morning‐nighttime densities at 210 km. Additionally, the two hump structure in the middle

A12306

thermosphere density is mainly associated with the tides that are excited above 30 km.

5. Summary [15] Observations from the SETA3 satellite during 1 September 1983 to 31 October 1983 were used to examine the variations of lower thermosphere density at daytime (1030 LT) and nighttime (2230 LT). As the measured densities at satellite altitudes are referenced to an altitude at 210 km, both daytime and nighttime densities showed an obvious double hump structure in the equatorial region. Specifically, at 1030 LT, there are two crests occurring at around ±20°. At 2230 LT, the northern crest locates within the latitudinal range of 20–40°N, whereas the southern crest appears at 40–50°S. Moreover, the equatorial density trough appears to correspond to the geographic rather than geomagnetic equator in both daytime and nighttime sectors. The crest‐to‐trough ratios from the observations are around 1.05 and 1.15 for 1030 and 2230 LT, respectively. Our results are different from those of Ching and Carter [1974], who showed an equatorial maximum in nighttime (about 0130 LT) mass density organized by the geomagnetic equator. Finally, the TIMEGCM simulation was conducted to understand the mechanism for latitudinal structure in the middle thermosphere density. The simulations reveal that the observed two hump structures in both daytime and nighttime sectors are mainly associated with the tides excited above 30 km, although they are modulated by the tides originating from the troposphere and below. [16] Acknowledgments. This work was supported by the National Natural Science Foundation of China (41174139, 41121003, 41025016, 40890165). The Project of Chinese Academy of Sciences KZCX2‐EW‐ QN509, and Thousand Young Talents Program. J.F. was supported under AFOSR MURI award FA9550‐07‐1‐0565 to the University of Colorado at Boulder. H.‐L.L. wants to acknowledge NSF CEDAR grant ATM‐ 0836386. The National Center for Atmospheric Research is sponsored by the National Science Foundation. J.L. thanks Astrid Maute for her great help in this study. [17] Robert Lysak thanks the reviewers for their assistance in evaluating this manuscript.

References Akmaev, R. A., F. Wu, T. J. Fuller‐Rowell, and H. Wang (2009), Midnight temperature maximum (MTM) in Whole Atmosphere Model (WAM) simulations, Geophys. Res. Lett., 36, L07108, 4 pp., doi:10.1029/ 2009GL037759. Akmaev, R. A., F. Wu, T. J. Fuller‐Rowell, H. Wang, and M. D. Iredell (2010), Midnight density and temperature maxima, and thermospheric dynamics in Whole Atmosphere Model (WAM) simulations, J. Geophys. Res., 115, A08326, 8 pp., doi:10.1029/2010JA015651. Arduini, C., G. Laneve, and F. A. Herrero (1997), Local time and altitude variation of equatorial thermosphere midnight density maximum (MDM): San Marco drag balance measurements, Geophys. Res. Lett., 24, 377–380, doi:10.1029/97GL00189. Bowman, B. R., W. K. Tobiska, F. A. Marcos, and C. Valladares (2008), The JB2006 empirical thermospheric density model, J. Atmos. Sol. Terr. Phys., 70(5), 774–793, doi:10.1016/j.jastp.2007.10.002. Ching, B. K., and V. L. Carter (1974), Ion gauge measurements of latitudinal density variations at night, Geophys. Res. Lett., 1(2), 93–96, doi:10.1029/GL001i002p00093. Forbes, J., F. Marcos, and F. Kamalabadi (1995), Wave structures in lower thermosphere density from satellite electrostatic triaxial accelerometer me asu rem ents, J. Ge ophy s. Re s., 100( A8), 14, 693–1 4,701, doi:10.1029/95JA00065. Hagan, M. E., and J. M. Forbes (2002), Migrating and nonmigrating diurnal tides in the middle and upper atmosphere excited by tropospheric latent

6 of 7

A12306


heat release, J. Geophys. Res., 107(D24), 4754, doi:10.1029/ 2001JD001236. Hagan, M. E., and J. M. Forbes (2003), Migrating and nonmigrating semidiurnal tides in the upper atmosphere excited by tropospheric latent heat release, J. Geophys. Res., 108(A2), 1062, doi:10.1029/2002JA009466. Hagan, M. E., A. Maute, R. G. Roble, A. D. Richmond, T. J. Immel, and S. L. England (2007), Connections between deep tropical clouds and the Earth’s ionosphere, Geophys. Res. Lett., 34, L20109, doi:10.1029/ 2007GL030142. Hedin, A. E., and H. G. Mayr (1973), Magnetic control of the near equatorial neutral thermosphere, J. Geophys. Res., 78, 1688–1691, doi:10.1029/JA078i010p01688. Herrero, F. A., and N. W. Spencer (1982), On the horizontal distribution of the equatorial thermospheric midnight temperature maximum and its seasonal variation, Geophys. Res. Lett., 9, 1179–1182, doi:10.1029/ GL009i010p01179. Lei, J., J. P. Thayer, and J. M. Forbes (2010), Longitudinal and geomagnetic activity modulation of the equatorial thermosphere anomaly, J. Geophys. Res., 115, A08311, doi:10.1029/2009JA015177. Liu, H., H. Lühr, V. Henize, and W. Köhler (2005), Global distribution of the thermospheric total mass density derived from CHAMP, J. Geophys. Res., 110, A04301, doi:10.1029/2004JA010741. Liu, H., H. Lühr, and S. Watanabe (2007), Climatology of the equatorial thermospheric mass density anomaly, J. Geophys. Res., 112, A05305, doi:10.1029/2006JA012199. Ma, R., J. Xu, W. Wang, J. Lei, H.‐L. Liu, A. Maute, and M. E. Hagan (2010), Variations of the nighttime thermospheric mass density at low and middle latitudes, J. Geophys. Res., 115, A12301, doi:10.1029/ 2010JA015784.

A12306

Philbrick, C. R., and J. P. McIsaac (1972), Measurements of atmospheric composition near 400 km, Space Res., 12, 743–750. Picone, J. M., A. E. Hedin, D. P. Drob, and A. C. Aikin (2002), NRLMSISE‐00 empirical model of the atmosphere: Statistical comparisons and scientific issues, J. Geophys. Res., 107(A12), 1468, doi:10.1029/2002JA009430. Raghavarao, R., L. E. Wharton, N. W. Spencer, H. G. Mayr, and L. H. Brace (1991), An equatorial temperature and wind anomaly (ETWA), Geophys. Res. Lett., 18, 1193–1196, doi:10.1029/91GL01561. Raghavarao, R., W. R. Hoegy, N. W. Spencer, and L. E. Wharton (1993), Neutral temperature anomaly in the equatorial thermosphere: A source of vertical winds, Geophys. Res. Lett., 20, 1023–1026, doi:10.1029/ 93GL01253. Rhoden, E. A., J. M. Forbes, and F. A. Marcos (2000), The influence of geomagnetic and solar variabilities on lower thermosphere density, J. Atmos. Sol. Terr. Phys., 62, 999–1013, doi:10.1016/S1364-6826(00)00066-3. Roble, R. G., and E. C. Ridley (1994), A thermosphere‐ionosphere‐ mesosphere‐electrodynamics general circulation model (TIME‐GCM): Equinox solar cycle minimum simulations (30–500 km), Geophys. Res. Lett., 21, 417–420, doi:10.1029/93GL03391. X. Dou, J. Lei, T. Li, X. Luan, and X. Xue, School of Earth and Space Sciences, University of Science and Technology of China, Anhui, Hefei 230026, China. ([email protected]; [email protected]) J. M. Forbes, Department of Aerospace Engineering Sciences, University of Colorado at Boulder, Boulder, CO 80309, USA. H.‐L. Liu, High Altitude Observatory, National Center for Atmospheric Research, PO Box 3000, Boulder, CO 80307, USA.

7 of 7