Equatorial and Low Latitude Ionospheric Effects During Sudden

Space Sci Rev DOI 10.1007/s11214-011-9797-5

Equatorial and Low Latitude Ionospheric Effects During Sudden Stratospheric Warming Events Ionospheric Effects During SSW Events Jorge L. Chau · Larisa P. Goncharenko · Bela G. Fejer · Han-Li Liu

Received: 21 March 2011 / Accepted: 6 June 2011 © Springer Science+Business Media B.V. 2011

Abstract There are several external sources of ionospheric forcing, including these are solar wind-magnetospheric processes and lower atmospheric winds and waves. In this work we review the observed ion-neutral coupling effects at equatorial and low latitudes during large meteorological events called sudden stratospheric warming (SSW). Research in this direction has been accelerated in recent years mainly due to: (1) extensive observing campaigns, and (2) solar minimum conditions. The former has been instrumental to capture the events before, during, and after the peak SSW temperatures and wind perturbations. The latter has permitted a reduced forcing contribution from solar wind-magnetospheric processes. The main ionospheric effects are clearly observed in the zonal electric fields (or vertical E × B drifts), total electron content, and electron and neutral densities. We include results from different ground- and satellite-based observations, covering different longitudes and years. We also present and discuss the modeling efforts that support most of the observations. Given that SSW can be forecasted with a few days in advance, there is potential for using the connection with the ionosphere for forecasting the occurrence and evolution of electrodynamic perturbations at low latitudes, and sometimes also mid latitudes, during arctic winter warmings. Keywords Atmosphere-ionosphere coupling · Equatorial aeronomy · Low latitude electrodynamics · Stratospheric warming J.L. Chau (!) Radio Observatorio de Jicamarca, Instituto Geofisico del Peru, Lima, Peru e-mail: [email protected] L.P. Goncharenko Massachusetts Institute of Technology, Haystack Observatory, Westford, MA, USA B.G. Fejer Center for Atmospheric and Space Sciences, Utah State University, Logan, UT, USA H.-L. Liu High Altitude Observatory, National Center for Atmospheric Research, Boulder, CO, USA

J.L. Chau et al.

1 Introduction and Scope Forecasting the ionospheric weather is very challenging. A significant portion of the day-today variability of the ionospheric parameters cannot be explained by the major drivers coming from above, i.e., solar ionizing flux and geomagnetic activity. The sources of such variability have been associated for many years with lower atmospheric processes (e.g., Rishbeth 2006). However the understanding of these links has been difficult due to the rapid response of the ionosphere to solar and magnetospheric drivers, and previously limited databases. In this work we review the systematic studies of one of these links, i.e., ionospheric variability during sudden stratospheric warming (SSW) events and in particular at equatorial and low latitudes. A SSW is a large meteorological event where westerly winds in the Northern winter stratosphere abruptly (i.e., in a few days time) slow down or even reverses direction, accompanied by a break-down of the polar vortex and a rise of stratospheric temperature by several tens of degrees. As a large-scale meteorological process in the winter polar region, it strongly affects vertical coupling in a large range of altitudes, from the stratosphere to the thermosphere (Liu and Roble 2002). It has been under extensive observational and theoretical investigation for many years (e.g., Andrews et al. 1987). The key mechanism for SSW events is the growth of planetary waves propagating upward from the troposphere and their non-linear interaction with the zonal mean flow (Matsuno 1971). The beauty of this connection is that SSWs are predicted a few days in advance (e.g., Kim and Flatau 2010), and therefore understanding this connection could help in forecasting the ionospheric weather. Research studying the lower atmosphere-ionosphere coupling during SSWs has been catalyzed after the observations reported by Goncharenko and Zhang (2008) and Chau et al. (2009). They clearly showed that the ion temperature at mid-latitudes and vertical E × B plasma drifts at equatorial latitudes, respectively, were correlated with the January 2008 SSW event. Those observations and results were a consequence of an unprecedented observing campaigns using all the incoherent scatter radars and other ground-based instruments. Since then, a significant number of upper atmospheric effects have been associated with this coupling, e.g.: – Multiple studies have demonstrated SSW effects in various ionospheric parameters: total electron content (TEC) (e.g., Goncharenko et al. 2010a, 2010b; Pedatella and Forbes 2010a; Chau et al. 2010), peak electron density (fo F2 ) (e.g., Yue et al. 2010; Pancheva and Mukhtarov 2011), electric fields (e.g., Anderson and Araujo-Pradere 2010; Fejer et al. 2010). – Differences in longitudinal response have been documented with observations over the Philippines (e.g., Anderson and Araujo-Pradere 2010; Fejer et al. 2010) and Indian (Vineeth et al. 2009; Sridharan et al. 2009) sectors. – Global character of the observed phenomena have been studied with satellite observations using the Challenging Minisatellite Payload (CHAMP) (e.g., Fejer et al. 2010), the Constellation Observing System for Meteorology, Ionosphere, and Climate (COSMIC) (Yue et al. 2010; Pancheva and Mukhtarov 2011). – The coupling mechanisms were investigated with modeling studies using the NCAR Thermosphere–Ionosphere–Mesosphere Electrodynamics General Circulation Model (TIMEGCM) (e.g., Liu et al. 2010) and the Whole Atmosphere Model (WAM) (e.g., Fuller-Rowell et al. 2010). Our paper is organized as follows. We first start with background material related to: (a) the ionospheric variability (Sect. 2), (b) the quiet-time equatorial vertical E × B drifts including their statistics and measurement techniques (Sect. 3), and (c) sudden stratospheric

Equatorial and Low Latitude Ionospheric Effects During Sudden

warming events (Sect. 4). We put special emphasis on the vertical E × B drifts since, as we will explain later, it is one of the most important parameters that controls the electrodynamics at equatorial and low latitudes. Then, we review the effects on ionospheric parameters that have been associated with the SSWs, in particular, the vertical E × B drifts, TEC, and fo F2 (Sect. 5). Finally, we discuss the possible scenarios including those supported by current numerical modeling efforts. 2 Ionospheric Variability There is a very good understanding of the F-region ionosphere climatology, i.e., mean values for different seasons, solar conditions, and degrees of magnetic activity, as should be expected after more than 70 years of ionospheric observations. However, the understanding of its weather and thus its predictive capability is very poor. This variability displays oscillations over a wide range of timescales (from minutes to severals days). During geomagnetic storms and substorms, there are significant ionospheric changes which can last for hours and it can take days for the ionosphere to recover. Even during nonstorm periods, geomagnetic perturbations associated with solar coronal holes and high-speed solar wind streams can cause periodic ionosphere and thermosphere oscillations with the solar rotation period and their subharmonics of this period (Lei et al. 2008). Ionospheric variability has been also correlated on a statistical basis with the Kp or Ap index (e.g., Rishbeth and Mendillo 2001; Pancheva et al. 2002; Altadill and Apostolov 2001; Xiong et al. 2006). During quiet times, i.e., under geomagnetically undisturbed conditions, significant ionospheric day-to-day variability is also observed. For example, it has long been recognized that there is large day-to-day variability in the vertical plasma drift observed at the magnetic equator, with this variability being the largest during solar minimum conditions (e.g., Fejer et al. 1991; Fejer and Scherliess 2001). Similarly, changes of up to 20% in daytime F-region maximum electron density that cannot be associated with solar or magnetic drivers have been reported by Rishbeth and Mendillo (2001) and Mendillo et al. (2002). These variations are comparable to ionospheric changes related to geomagnetic activity (e.g., Fuller-Rowell et al. 2000; Rishbeth and Mendillo 2001; Forbes et al. 2000). In the equatorial ionosphere, the monthly ionospheric variability (i.e., monthly standard deviation over monthly mean) reaches 30–40% at sunrise and 15–30% at sunset (e.g., Bilitza et al. 2004), while quiet-time variations in daytime electron density can be as high as 200% (Zhao et al. 2008). Mikhailov et al. (2007) also showed that there exist synchronous quiet day electron density variations in both the E and F regions. Recent observations by C/NOFS have revealed very large equatorial ionospheric depletion near dawn under geomagnetically quiet conditions (De la Beaujardière et al. 2009). The electrodynamic plasma (E × B) drifts, and in particular the vertical component, play important roles on the plasma distribution and dynamics in the low latitude ionosphere. In the next section (Sect. 3), we summarize the main characteristics of these drifts under geomagnetically quiet-time conditions. These characteristics will be later put into context with the observations under similar quiet-time conditions but during SSW periods. 3 Quiet-Time Equatorial Vertical E × B Drifts The quiet-time equatorial E × B drifts are driven by neutral wind generated E-region dynamo electric fields (e.g., Richmond et al. 1976; Stening 1981) and by F -region polarization electric fields (e.g., Rishbeth 1971; Heelis et al. 1974). The thermospheric winds are

J.L. Chau et al.

highly variable as a result of changes in the global tidal forcing, and effects of irregular winds, planetary, and gravity waves. Previous studies suggest that planetary waves and tides play important roles in the electrodynamics of the lower thermosphere (e.g., Chen 1992; Forbes and Leveroni 1992; Parish et al. 1994; Immel et al. 2009). In addition, the relative efficiencies of the E- and F -region dynamos change significantly throughout the day, and also with season and solar activity. For a comprehensive review on the physical principles underlying the electrodynamic properties of the equatorial and low-latitude ionosphere, we recommend reading Heelis (2004). Moreover, Fejer (2010) has recently reviewed the ionospheric electrodynamics at low latitudes. In this section we focus on the statistics and measurement techniques for the equatorial vertical E × B drifts. 3.1 Equatorial Vertical E × B Drift Statistics The equatorial vertical E × B plasma drifts have been extensively studied using groundbased instruments, i.e., coherent and incoherent scatter radars, ionosondes and magnetometers. These ground-based measurements started in 1968 using the Jicamarca Incoherent Scatter radar (Woodman and Hagfors 1969; Woodman 1970). The characteristics of the Jicamarca quiet-time vertical drifts have been described in several publications (e.g., Fejer et al. 1991; Scherliess and Fejer 1999; Fejer and Scherliess 2001). Figure 1 shows scatter plots of these drifts for periods around equinox and June solstice for two ranges of solar flux index. The December solstice results are not shown since there were not enough measurements for the period studied, i.e., between 1968 and 1992. These drifts show a large day-to-day variability at all local times. They are typically upward during the day (∼20 m/s) and downward at night (∼10–40 m/s) and do not change much with season and solar flux, except near sunset and sunrise. Amplitudes of the pre-reversal enhancement (around sunset) increase strongly with solar activity (e.g., Scherliess and Fejer 1999). The variability of these vertical drifts decreases from low to high solar flux periods for all seasons. The quiet-time variability of the Jicamarca vertical drifts is local time, seasonal, and solar cycle dependent. Figure 2 shows the mean values and standard deviations of these drifts for morning and afternoon periods as function of day of the year and two solar flux indexes. This variability is largest in the dawn–noon sector and during March equinox solar minimum periods, when the midday average upward drift velocity from consecutive magnetically quiet days can often change by more than 10 m/s. The day-to-day variability of the vertical drifts decreases in the afternoon sector and with the increase of solar activity for all seasons. For the results we present in later sections, it is important to take note of the December solstice daytime values, in particular: (1) the standard deviation is larger at lower solar flux indexes with a value of ∼7 m/s, and (2) the afternoon mean value is usually downward, particularly at low solar flux values. Combining Jicamarca vertical drifts and AE-E (Atmosphere-Explorer 2) satellite measurements, Scherliess and Fejer (1999) developed a quiet-time empirical F region vertical drift global model for different seasons and solar conditions. This model showed large longitudinal variations, particularly, for the evening upward drifts. Using drift measurements from the Republic of China Satellite (ROCSAT-1), Fejer et al. (2008) has developed a new empirical global model for moderate and high solar conditions for three seasons. They showed that the longitudinal dependence of the daytime and nighttime vertical drifts is much stronger than reported earlier, especially during December and June solstice. Among other important results, the morning and afternoon December solstice drifts have significantly different longitudinal dependence, and the daytime upward drifts have strong wave number-four signatures during equinox and June solstice. Using the same ROCSAT-1 data, Kil et al. (2009)

Equatorial and Low Latitude Ionospheric Effects During Sudden Fig. 1 Scatter plots of quiet-time Jicamarca F -region vertical plasma drifts (positive upward) during equinox (March-April, September-October) and June solstice (May-August) for two ranges of the solar flux index (Φ) (from Fejer and Scherliess 2001, Fig. 1)

Fig. 2 Midday and afternoon quiet-time vertical plasma drift velocities for two ranges of the solar flux index (top panels). Standard deviations of the vertical drifts (bottom panels) (from Fejer and Scherliess 2001, Fig. 2)

introduced a similar empirical drift model but for every month instead of three seasons. Regional models at other longitude sectors have been also developed using ground-based ionosonde measurements in the Brazilian (Batista et al. 1996) and Indian (Sastri 1996) sectors.

J.L. Chau et al.

Fig. 3 Range-time plot of vertical drifts measured at Jicamarca on April 10, 1997. Red indicates upward drift while blue indicates downward drift (from Kudeki et al. 1999, Plate 1)

Other indirect measurements of daytime vertical F region drifts (or zonal electric fields) at different longitudes, particularly at equatorial latitudes, have been provided by the CHAMP satellite using magnetic field measurements of the equatorial electrojet (Lühr et al. 2008). These measurements show a strong longitudinal modulation of the equatorial daytime eastward electric fields associated with the wavenumber-four structure in agreement with the ROCSAT-1 measurements. These ionospheric signatures have been identified before in measurements of the nighttime equatorial ionization anomaly (e.g., Immel et al. 2006) and noontime equatorial electrojet (England et al. 2006) using far ultraviolet images. The reported wavenumber-four structure represents another evidence of atmosphere-ionosphere coupling, but this connections is not part of this review. 3.2 Equatorial Vertical E × B Measurements Most of the equatorial drift measurements have been made with the Jicamarca Incoherent Scatter radar (ISR). Here we describe those measurements and also other ground-based measurements with a particular emphasis on the daytime capabilities. As we will see later, so far most of the ionospheric effects associated with the SSW events have been identified during the day. Since 1968 (Woodman and Hagfors 1969), very precise measurements of vertical plasma drifts have been made at the Jicamarca Radio Observatory (JRO) from ISR echoes. By using antenna beams pointing perpendicular to the magnetic field, the correlation times of the received signals are very long, allowing the use of pulse-to-pulse techniques. From 1968 to 1994, daytime and nighttime measurements with 1–2 m s−1 precision have been obtained few days per year. Using a spectral analysis and inversion techniques, Kudeki et al. (1999) introduced a new mode that improved the precision further and allowed higher altitude resolution. Measurements using this new approach started in 1994. Figure 3 shows a typical example of the type of vertical drifts that are measured at Jicamarca. Those measurements are obtained with an altitude resolution of 15 km and time resolution that varies between 1 and 5 min. The altitudinal coverage varies depending on the electron density profile. Since 2001 equatorial daytime vertical drifts have been routinely derived from the Doppler shift measurements of coherent echoes occurring in the equatorial F1 region, the socalled 150-km echoes (Kudeki and Fawcett 1993). Although the physics behind these echoes is not known yet, their vertical velocities are in excellent agreement with the F region vertical drifts (Chau and Woodman 2004). Given that 150-km echoes are much stronger than incoherent scatter echoes, such measurements are done with the JULIA (Jicamarca Unattended Long-term Investigations of the Atmosphere) system that uses the large Jicamarca


Fig. 4 Example of 150-km parameters obtained with the JULIA radar on January 3, 2003: (a) Signal– to-noise-ratio (dB), (b) vertical drift (m/s), and (c) spectral width (m/s). The solid in the top panel represents the average (in range) drifts (from Chau and Kudeki 2006, Fig. 1)

array but low power transmitters, allowing a better time coverage (between 125–180 days per year). The precision of these velocities is comparable to the ISR drifts, however they provide only mean values for the F region. Figure 4 shows an example of typical radar parameters from 150-km echoes. These measurements are only possible during the day and in the case of Jicamarca, the echoes do not show a strong seasonal or solar dependence (e.g., Chau and Kudeki 2006). Similar measurements are now routinely made at other longitudinal sectors: Brazil (de Paula and Hysell 2004), India (Patra and Rao 2007), and Pohnpei (Tsunoda and Ecklund 2004). It should be noted that the 150-km and F region vertical drifts are quite similar but not identical, otherwise there would be a violation of the curl-free conditions. The vertical drifts increase with height in the morning and decrease in the afternoon as shown by Pingree and Fejer (1987). Recently, Alken (2009) presented a model of vertical plasma drifts from 150-km echoes observed by the JULIA system. The model includes a climatology of the 150 km equatorial vertical drifts as well as an estimate of the day-to-day variability, which can be significant, for different seasons, solar conditions and daytime (0800 to 1600 LT) hours in the Peruvian sector. Figure 5 shows the drift values as function of day of the year and daytime hour for low solar activity conditions: (a) the Scherliess/Fejer model (Scherliess and Fejer 1999), (b) the JULIA data, and (c) the JULIA Vertical Drift model (JVDM). There is very good agreement between the JVDM model and the data. All major features of the vertical plasma drifts have been reproduced. The drift maxima during March and September equinox and minima during June and December solstice agree well. An important improvement in the JVDM compared to the Scherliess and Fejer model is the allowance of differences between March and September equinox. Again, for the purpose of this review, the most important results are the December solstice values, i.e., positive values in the morning and negative values in the afternoon.

J.L. Chau et al.

Fig. 5 Vertical drifts values for low solar conditions: (left) Scherliess and Fejer model output, (middle) raw JULIA vertical drift data, and (right) JULIA Vertical Drift Model (from Alken 2009, Fig. 4)

Equatorial daytime vertical drifts are also now routinely derived from the difference of horizontal geomagnetic fields measured at pairs of stations, one very near the magnetic equator and another a few degrees off at about the same longitude (e.g., Anderson et al. 2002). Vertical drifts derived using magnetic field data from Jicamarca and Piura (∼6.8◦ magnetic) are generally in good agreement with radar measured drifts from both 150-km echoes and Incoherent scatter signals, but their accuracy can sometimes be affected by magnetic field changes produced by variable low latitude E region winds (e.g., Fang et al. 2008) or unexpected changes in the E region conductivity (e.g., during solar flares). Anderson et al. (2006) has used this technique at other longitudes. In Fig. 6 we show examples of daytime vertical E × B drifts using (a) ISR, (b)150-km echoes, and (c) magnetometer ∆H, at Jicamarca for similar dates. The bottom panels shows the values during the January 2009 event, and the top panel shows the control days (similar season and solar conditions but without SSW) for all three techniques. In general, all three techniques are in very good agreement, particularly around noontime hours. In Table 1 we summarize the main parameters for all three techniques. Recently, equatorial zonal electric fields have been derived from equatorial magnetic field measurements by the CHAMP satellite (Alken and Maus 2010). Although we are focusing on daytime vertical drift measurements, evening and early night equatorial F region vertical plasma drifts are routinely derived from ionosonde observations (e.g., Abdu et al. 2007). The accuracy of the ionosonde derived nighttime vertical drifts is improved when production and recombination effects are properly taken into account using theoretical models (e.g., Bertoni et al. 2006). On the other hand the ionosonde derived vertical velocity during the day does not represent the F region vertical E × B drift, particularly at the equator (Woodman et al. 2006). 4 Sudden Stratospheric Warming Events In this section we describe the large SSW meteorological phenomena. Then, we summarize some of the effects that are not part of this review, i.e., they occur outside the equatorial/low-


Fig. 6 Examples of daytime E × B vertical drifts using: (a) ISR, (b) 150-km echoes, (c) ∆H. The bottom panels show days (in different colors) around the Jan 2009 SSW event, and top panels show control days for a similar season and solar conditions. The black solid line and the dashed lines represent the expected average and standard deviations values, respectively. (Adapted from Chau et al. 2009, Fig. 2) Table 1 E × B vertical drift measurements at Jicamarca using different techniques: (a) ISR, (b) 150-km echoes, (c) Magnetometer ∆H Value

ISR

150-km echoes

∆H Mag. at JRO and Piura

Instrument

Jicamarca ISR

JULIA

Annual coverage

∼20 days

>150 days

>300 days

Since

1968

2001

2000

Time coverage

All day*

07-18 LT

09-17 LT

Time res.

5 min

1–5 min

1 min

Accuracy