El Niño–Southern Oscillation effect on ionospheric tidal/SPW amplitude in 2007–2015 FORMOSAT-3/COSMIC observations
- 305 Downloads
KeywordsEl Niño–Southern Oscillation Ionospheric tide/SPW Quasi-biennial oscillation Radio occultation
El Niño–Southern Oscillation
total electron content
diurnal eastward wavenumber 3
stationary planetary wavenumber 4
diurnal westward wavenumber 1
semidiurnal westward 2
mesosphere and lower thermosphere
equatorial ionospheric anomaly
Thermosphere Ionosphere Mesosphere Energetics and Dynamics
TIMED Doppler Interferometer
Global Positioning System
low earth orbit
Oceanic Niño Index
Ionospheric plasma density structures are highly variable on broad range of temporal and spatial scales, and the solar activity and lower atmosphere forcing are the major drivers. The El Niño–Southern Oscillation (ENSO) ocean–atmosphere coupling phenomenon is one of the significant sources of variability in the thermosphere (Gurubaran et al. 2005; Lieberman et al. 2007; Pedatella and Liu 2012, 2013; Warner and Oberheide 2014; Liu 2016; Liu et al. 2017; Sun et al. 2018). However, studies on the ENSO signature in the ionosphere are rare and challenging (Pedatella and Forbes 2009; Pedatella and Liu 2013; Chang et al. 2018) due to the complex nature of the atmosphere–ionosphere-solar system. Pedatella and Liu (2013) simulated the influence of the ENSO-driven tidal variability on the low-latitude ionosphere. The simulations show that the interannual tidal variability in the mesosphere and lower thermosphere (MLT) can introduce 10–15% variability in the E × B vertical drift velocity and ionosphere peak density. Chang et al. (2018) analyzed the FORMOSAT-3/COSMIC (F3/C) S4 index from 2007 to 2014 and showed that ENSO signatures can transmit to Es formation mechanisms, potentially through modulation of vertically propagating atmospheric tides that alter lower thermospheric wind shears.
The ENSO-related variation in diurnal tides in the MLT (Gurubaran et al. 2005; Lieberman et al. 2007; Pedatella and Liu 2012, 2013; Warner and Oberheide 2014; Liu et al. 2017; Sun et al. 2018) can modulate the ionospheric wind dynamo which drives the equatorial ionospheric anomaly (EIA) wave-4 longitudinal structure on both sides of the magnetic equator via the fountain effect (Lühr et al. 2007; Wan et al. 2010; Fang et al. 2013). The diurnal eastward wavenumber 3 (DE3) and stationary planetary wavenumber 4 (SPW4) in the ionospheric electron density comprise the largest portion of the wave-4 structure (Pancheva and Mukhtarov 2012; Chang et al. 2013, 2016). Therefore, this study analyzes the total electron content (TEC) of F3/C to explore the ENSO effect on the ionospheric DE3 and SPW4 amplitudes from January 2007 to December 2015. The period comprises two ENSO warm (2009/2010, 2014/2015)/cold (2007/2008, 2010/2011) phases. The rest of the periods are weak ENSO phases and neutral conditions (according to the Ocean Niño Index, http://www.cpc.ncep.noaa.gov/products/analysis_monitoring/ensostuff/ensoyears.shtml).
Warner and Oberheide (2014) reported the notable increases in the wind DE3 amplitude in the lower thermosphere within the 0◦–20◦N latitude band during the ENSO cold phase (2007/2008 and 2010/2011). Especially during the 2010/2011 strong ENSO cold phase, enhanced forcing from below during the winter months causes a ~ 70% increase in the zonal wind DE3 tide in the MLT region. They also suggested that constructive interference of the antisymmetric and symmetric tidal contributions results in the larger MLT zonal wind DE3 amplitude in the Northern Hemisphere during the strong La Niña. Accordingly, this study also conducts the zonal wind DE3 amplitude observed by Thermosphere Ionosphere Mesosphere Energetics and Dynamics (TIMED)/TIMED Doppler Interferometer (TIDI) at 100 km altitude to suggest a causal mechanism for the ionospheric tide/SPW response to the ENSO phases. Note that the ionospheric diurnal westward wavenumber 1 (DW1) primarily reflects in situ photoionization that produces ionization peaks. The numerical experiments by Hagan et al. (2001) suggested that the MLT DW1 cannot vertically propagate above 130 km altitude. Comparing the TEC DW1 with the other components is helpful to recognize the TEC tide/SPW response to the ENSO-related variation from the lower thermosphere.
Moreover, the quasi-biennial oscillation (QBO) is an important and persistent feature in the thermosphere and ionosphere. The ENSO can modulate the stratospheric QBO (Taguchi 2010) as well as the QBO components in the temperature DW1 and DE3 in the MLT (Sun et al. 2018). Both the wind DE3 in the lower thermosphere and solar activity can affect the QBO in the TEC tidal/SPW amplitudes (Chang et al. 2016; Wang et al. 2018). However, it is challenging to distinguish the ionospheric QBO response to the stratospheric QBO and the QBO component in solar activity because they yield similar periodicities (Tang et al. 2014). Accordingly, we examine whether the ENSO is capable of influencing the QBO in the ionospheric tide/SPW.
Data and methodology
The Global Positioning System (GPS) radio occultation (RO) experiment onboard the F3/C satellites observed ionospheric electron density (from 70 to 800 km altitude) for over 10 years since their launch in 2006. The F3/C mission consists of six microsatellites in the different mission planes and provides observations of vertical structures of the ionospheric electron density globally by using the radio occultation observation technique. The electron density profiles from F3/C have significantly improved our understanding of the global morphology and variability of the dynamical EIA structures (Lin et al. 2007, 2012; Maruyama et al. 2016). The F3/C electron density profiles are adopted from the ionPrf file in the second data level which is processed by the COSMIC Data Analysis and Archive Center (http://cosmic-io.cosmic.ucar.edu/cdaac/index.html) using the Abel inversion technique from TEC along LEO (low earth orbit)—GPS rays since May 2006. Yue et al. (2010) and Liu et al. (2010) reported that the electron density profiles obtained by the Abel inversion have large amount of error below 200 km altitude. The F3/C TEC used in this study is the integration of electron density from 200 km to the LEO satellite altitude (800 km). The global TEC observation of F3/C RO with uniformly spatial and temporal distributions from 2007 to 2015 is essential for properly resolving the tidal/SPW components with higher longitudinal wavenumber. The sounding rate of the F3/C RO is low after 2015.
The zonal wind in the lower thermosphere was obtained by the Thermosphere Ionosphere Mesosphere Energetics and Dynamics (TIMED)/TIMED Doppler Interferometer (TIDI) from January 2007 to December 2015 (http://download.hao.ucar.edu/archive/tidi/data/vec0307a/). The TIDI wind observation is successfully utilized in studying migrating and non-migrating tides in the MLT (Wu et al. 2008a, b).
In the left panels of Fig. 4, the components with the period interval of 0–12 months are highly fluctuating through the entire period from 2007 to 2015. It is difficult to glean the ENSO signature from the short-period components (0–12-month period band) because they are highly fluctuated. The middle panels show that the components of the TEC DW1 and F10.7 with the period interval of 12–24 months agree well from 2011 to 2015 (higher solar activity period). The components of DW1 and F10.7 are out of phase to the wind DE3 near the 2010–2011 winter. It is due to the fact that the TEC DW1 is highly controlled by the solar activity. The wind DE3 in the lower thermosphere cannot control the DW1 but can induce a DE3 variation in TEC. The wind DE3 increases significantly near the 2010–2011 winter, the components of the wind DE3 and TEC DE3/SPW4 show good agreement during the 2010/2011 ENSO cold phase when the crests of the TEC DE3/SPW4 and wind DE3 are closer to each other. There is a difference of few months between the crests of the wind DE3 and TEC DE3/SPW4. The right panels show the comparison of the components with the period interval of 24–48 months. It is not surprising to see the good agreement between the long-term variation in the TEC DW1 amplitude anomaly and F10.7. On the other hand, it is hard to distinguish the effect of solar activity and wind DE3 on the long-term variation in the TEC DE3 and SPW4, because the few cycles yield similar periodicities and are nearly in phase with each other. In the Southern Hemisphere, the TEC DW1 and F10.7 agree well with each other in varies scales especially during the higher solar activity period. It is difficult to indicate the connection between the TEC DE3/SPW4 and MLT wind DE3 components with period intervals of 12–24 months and 24–48 months due to the contamination of solar activity and the weaker wind DE3 in the Southern Hemisphere (Fig. 3c). We do not show the results for the Southern Hemisphere because the amplitude anomalies with period longer than 1 year are weaker there.
The QBO components as shown in Fig. 5 agree with the QBOs decomposed from the TEC tidal/SPW amplitudes using the adaptive method named multi-dimensional ensemble empirical mode decomposition (Chang et al. 2016). Chang et al. (2016) reported that the TEC SPW4 and DE3 QBOs start increasing in middle 2010, but the DW1 and SW2 (semidiurnal westward 2) as well as the QBO component of F10.7 start increasing in early 2011. The band-pass filter applied in this study bases on the Fourier assumption, but the multi-dimensional ensemble empirical mode decomposition is an adaptive method without a basis function assumption. The results from the two independent methods agree with each other. The agreement reveals that, besides the effect of the solar radiation and ionosphere/thermosphere composition changes, the ENSO-related wind DE3 variation contributes to the QBO in the TEC DE3 and SPW4 during the strong 2010/2011 ENSO cold phase. The response of the wind DE3 in the lower thermosphere to the 2007/2008 ENSO cold phase is not as significant as that to the 2010/2011 ENSO cold phase (Warner and Oberheide 2014). It can be the reason of that the ionosphere response to the 2007/2008 ENSO cold phase is weak. Warner and Oberheide (2014) suggested that the ENSO-related enhanced forcing from below causes an increase in the zonal wind DE3 tide in the MLT region especially during the winter months of the 2010/2011 strong ENSO cold phase. Note that we have examined all the TEC tidal/SPW components with zonal wave number ranging from -4 to +4 and tidal harmonic from the diurnal to terdiurnal modes. The ENSO-related signature is most pronounced in the TEC DE3 and SPW4.
The stratospheric QBO can affect the diurnal tides in the lower thermosphere (Forbes et al. 2008; Oberheide et al. 2009) and ionosphere (Tang et al. 2014; Chang et al. 2016; Wang et al. 2018). The wind DE3 amplitude shows an intense peak with period shorter than 2 years during the 2010/2011 ENSO cold phase (Fig. 3b). However, the wavelet spectrum of the U30 index that stands for the stratospheric QBO does not show strong short-period signals during that period (Fig. 4 in Sun et al. 2018). Warner and Oberheide (2014) showed similar percent variations in DE3 heating and tidal wind deviations during the period of 2008–2011, indicating a strong ENSO influence, with a few exceptions that can be accounted for by the effects of the stratospheric QBO. Therefore, the QBO components of the TEC DE3 and SPW4 amplitudes are driven by the ENSO signature in the wind DE3, rather than the stratospheric QBO.
The series of ENSO warm phases shorten the QBO components in the zonal wind and temperature diurnal tides in the lower thermosphere from 2002 to 2007 (Sun et al. 2018). However, the individual ENSO warm phase of 2009/2010 may not allow us to show a robust TEC tidal/SPW response. This could be due to the negligible response of the wind DE3 during the ENSO warm phase (Warner and Oberheide 2014). Different types of ENSO warm phases, which are usually based on the location of the largest positive sea surface temperature anomalies, can result in different effects on the atmosphere (Johnson 2013).
The 2014/2015 ENSO warm phase is strong but occurs in the high-solar-activity period. The ENSO would impact ionosphere during the entire F3/C operational period, but the high solar activity contaminates the ENSO effect from the lower atmosphere. The neutral DE3 coupling into the ionosphere via the E-region dynamo can be modulated by the DW1 component of the conductivities and produce the SPW4 in electron density (Chang et al. 2013). The dynamo-driven ionospheric tidal/SPW is sensitive to the solar activity because the conductivities are highly dependent on the solar irradiance variation. Therefore, the contamination of solar activity can obscure the ENSO signature in the ionospheric tidal/SPW if the ENSO effect transmits into the ionosphere via the dynamo process. Longer data sets including numerous ENSO warm phases under low-solar activity conditions are required for studying the ionospheric tidal/SPW response to the ENSO warm phase.
The stratospheric QBO period is flexible and its mean period is near 28 months. However, the peak in the longer periods occurs close to 36 months in the solar activity that can border the QBO period range (18–34 months) of the ionospheric tide/SPW. The peaks of F10.7 and wind DE3 in the period near 1.5 years in 2010–2011 can also shorten the ionospheric QBO. The impact from the solar activity results in the period of the ionospheric QBO is more flexible than that of the QBO in the lower atmosphere.
On the short-period scales, the good agreement between the spectra of the F10.7 index (Fig. 3a) and the TEC DW1 amplitude anomalies (Fig. 2a, b) suggests that the solar activity induces the spectral peaks of the TEC DW1 with a period of ~ 0.6 years in both hemispheres in middle 2011. On the other hand, the spectral peaks of the TEC DE3 and SPW4 amplitude anomalies with period of ~ 0.3 years in early 2011 (Fig. 2c, e) may reflect the notable spectral peak of the wind DE3 amplitude anomaly with shorter periods in the Northern Hemisphere during the 2010/2011 ENSO cold phase (Fig. 3b). The spectral peak of the TEC DE3/SPW4 leads that of the F10.7 by nearly a half year. Therefore, the solar activity cannot induce the short-period signals of the TEC DE3/SPW4 amplitude in early 2011, but perturbs them in middle 2011.
Both the amplitude anomalies of the TEC DE3/SPW4 (Fig. 2c, e) and wind DE3 (Fig. 3a) strengthen in the Northern Hemisphere during the 2010/2011 ENSO cold phase. This reveals that some other causal mechanisms result in the anomalies appearing in the Northern Hemisphere besides the ionospheric dynamo, which drives the ionosphere in both hemispheres. Forbes et al. (2009) first showed that the upper thermospheric temperature DE3 represents the direct vertical propagation of non-migrating tides upward from the troposphere. Oberheide et al. (2009) applied a physics-based empirical fit model to connect the tides from TIMED satellite to the CHAMP satellite at 400 km. They found that the DE3 amplitudes in the upper thermosphere can increase by a factor of 3 in the zonal wind, by 60% in temperature and by a factor of 5 in density, caused by reduced dissipation above 120 km during solar minimum. It is therefore possible that the ENSO signature in the MLT neutral DE3 (Warner and Oberheide 2014) is capable of transmitting into the F region altitude in the Northern Hemisphere. The nonlinear interaction between the DE3 and in situ generated DW1 in the neutral thermosphere can produce an SPW4 (Hagan et al. 2009; Pedatella et al. 2012). The ENSO signature in DE3 and SPW4 in the upper thermosphere could transmit to the ionosphere through ion drag.
The ionospheric total electron content (TEC) diurnal eastward 3 (DE3) and stationary planetary wave 4 (SPW4) amplitudes measured by FORMOSAT-3/COSMIC respond to the ENSO signature in the lower thermophore during the lower-solar-activity period from 2007 to 2011. The 2010/2011 strong ENSO cold phase significantly enhances the TEC DE3 and SPW4 amplitudes with periods from 1 year to quasi-biennial periodicity at low latitude of the Northern Hemisphere. The QBO crests of the TEC DE3/SPW4 and the wind DE3 at lower thermosphere are almost in phase during the 2010/2011 ENSO cold phase. The solar activity dependence is hard to be entirely removed from the observation. Fixing the solar activity in the model simulation will enable extraction of ENSO signature uncontaminated by the solar activity.
YYS, HL, YM, LCC, and LBL designed and performed experiments. YYS analyzed data and drafted the manuscript. LCC developed the tidal fitting tool. All authors read and approved the final manuscript.
The authors declare that they have no competing interests.
Availability of data and materials
The FORMOSAT-3/COSMIC ionospheric electron density observation is downloaded from ftp://cdaac-ftp.cosmic.ucar.edu/. The TIMED/TIDI zonal wind observation is downloaded from http://download.hao.ucar.edu/archive/tidi/data/vec0307a/.
Consent for publication
Ethics approval and consent to participate
Y Y. S. was support by the NICT International Exchange Program. H. L. was support by JSPS KAKENHI grants 18H01270, 18H04446, and 17KK0095. Y. M. was supported by JSPS KAKENHI grant (B)15H03733. L. C. C. was supported by grants MOST 103-2111-M-008-019-MY3, 106-2111-M-008-010, and 107-2111-M-008-002-MY3 from the Taiwan Ministry of Science and Technology. L. B. L. was supported by National Natural Science Foundation of China (41621063).
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
- Lin CH, Wang W, Hagan ME, Hsiao CC, Immel TJ, Hsu ML, Liu JY, Paxton LJ, Fang TW, Liu CH (2007) Plausible effect of atmospheric tides on the equatorial ionosphere observed by the FORMOSAT-3/COSMIC: three-dimensional electron density structures. Geophys Res Lett 34:L11112. https://doi.org/10.1029/2007GL029265 CrossRefGoogle Scholar
- Maruyama N, Sun YY, Richards PG, Middlecoff J, Fang TW, Fuller-Rowell TJ, Akmaev RA, Liu JY, Valladares C (2016) A new source of the midlatitude ionospheric peak density structure revealed by a new Ionosphere-Plasmasphere model. Geophys Res Lett 43:2429–2435. https://doi.org/10.1002/2015GL067312 CrossRefGoogle Scholar
- Torrence C, Compo GP (1998) A practical guide to wavelet analysis. Bull Am Meteorol Soc 79(1):61–78. https://doi.org/10.1175/1520-0477(1998)079%3c0061:APGTWA%3e2.0.CO;2 CrossRefGoogle Scholar
- Wu DL, Hays PB, Wilbert RS (1995) A least-squares method for spectral-analysis of space-time series. J Atmos Sci 52(20):3501–3511. https://doi.org/10.1175/1520-0469(1995)052%3c3501:ALSMFS%3e2.0.CO;2 CrossRefGoogle Scholar
- Wu Q, Ortland DA, Killeen TL, Roble RG, Hagan ME, Liu HL, Solomon SC, Xu J, Skinner WR, Niciejewski RJ (2008a) Global distribution and interannual variations of mesospheric and lower thermospheric neutral wind diurnal tide: 1. Migrating tide. J Geophys Res 113(A5):A05308. https://doi.org/10.1029/2007JA012542 CrossRefGoogle Scholar
- Wu Q, Ortland DA, Killeen TL, Roble RG, Hagan ME, Liu HL, Solomon SC, Xu J, Skinner WR, Niciejewski RJ (2008b) Global distribution and interannual variations of mesospheric and lower thermospheric neutral wind diurnal tide: 2. Nonmigrating tide. J Geophys Res 113(A5):A05309. https://doi.org/10.1029/2007JA012543 CrossRefGoogle Scholar
- Xu J, Smith AK, Liu HL, Yuan W, Wu Q, Jiang G, Mlynczak MG, Russell JM III, Franke SJ (2009) Seasonal and quasi-biennial variations in the migrating diurnal tide observed by Thermosphere, Ionosphere, Mesosphere, Energetics and Dynamics (TIMED). J Geophys Res 114:D13107. https://doi.org/10.1029/2008JD011298 CrossRefGoogle Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.