International Journal of Geomagnetism and Aeronomy
Published by the American Geophysical Union
Vol. 1, No. 1, April 1998

The influence of the semidiurnal tide on altitude variations of the sporadic E layer

A. D. Akchurin, E. Yu. Zykov, N. A. Makarov, R. G. Minullin, Yu. A. Portnyagin, and O. N. Sherstyukov

Kazan' State University, Kazan', Tatarstan, Russia
1. Introduction
2. Initial Experimental Data
3. Method of Processing of Ionospheric and Wind Data
4. Results of Initial Data Processing
5. Discussion
6. Conclusion


Parameters of the vertical motion of the E region observed by ionosonde are compared with the wind parameters calculated by tidal and wind shear theories using meteor radar data. Simultaneous downward motions of the Es layers and the nodes of the semidiurnal tide are detected. From statistical analysis of the monthly data set, the influence of the complicated structure of the semidiurnal tide on the diurnal variation of the Es layers is revealed. The most probable time intervals of influence of the nodes of the (2, 2) mode and the modes of smaller vertical scale are determined.

1. Introduction

Chavdarov et al. [1975], Chimonas and Axford [1968], Fomichev and Shved [1981], Gershman et al. [1976], and Mathews and Bekeny [1979] have used wind shear theory (WST) to study the influence of the semidiurnal tide on the formation of the Es layer. However, coincidence of the expected time of occurrence of the Es layers by WST and the actual time of their appearance on the ionograms rarely occurs. The discrepancy may be accounted for by the longitudinal separation (>3000 km) between the regions where the Es layer parameters and the wind velocities in the meteor zone have been measured [Fomichev and Shved, 1981] and by the short duration of some measurements [Mathews and Bekeny, 1979].

In this paper we continue the investigation of tidal impact on the diurnal, vertical, and frequency variations of parameters of the midlatitude Es layer using data from closely located radar and ionospheric stations.

2. Initial Experimental Data

Temporal series of the values of the limiting frequency foEs and virtual height h'Es of the Es layer obtained by the "Tsyklon" ionosonde [ Minullin et al., 1994] with a vertical resolution of about 2.5 km are used in this paper. The measurements were conducted in July 1993 in the vicinity of Kazan (55oN, 49oE). The values of the real height hEs were calculated by the Mal'tseva et al. [1973] method. The Es traces were split into two types because of the difference in their formation. The first one is conventionally called the "main" type. In the daytime and nighttime it is usually represented by the "c" and "f" type layers, respectively. When these layers were absent in the ionogram, an Es layer of any other type (for example, of the "l" type) was accepted as a "main" layer. The Es layers which could not be considered as "main" (if, for example, there are two layers in the ionogram) were referred to as an "accompanying" type. Both the consequent Es [Robinson, 1960] (the ionogram trace of which has the shape of a small crescent and is situated between the E and F traces) and the additional Es [Chavdarov, 1975] (the upper one out of two traces observed in the ionogram) belong to this type. Cases in which Es of the "l" type appeared in the ionogram when the main trace was present were classified as "accompanying." Multiple reflections were not considered.

The database of velocity components of the neutral wind in the meteor zone resulted from measurements made by the meteor radar in Obninsk (55oN, 38oE).

The distance between the regions observed by the ionosonde and the radar is about 500 km and is less than the spatial scale of pulsations of the parameters of the semidiurnal tide (~1500 km [Portnyagin, 1981]).

3. Method of Processing of Ionospheric and Wind Data

According to WST the positive long-lived ions should accumulate in the height interval 95-120 km at the altitude above which the neutral wind is westward (so the ions entrained by it move downward) and below which the wind is eastward (the ions move upward) [Chimonis and Axford, 1968; Gershman et al., 1976]. Such a reversal of the vertical profile of the zonal wind is called a converging node of the zonal wind [ Chimonas and Axford, 1968; Whitehead, 1989]. A reversal with the opposite combination (the eastward wind is located above the westward one), where the electron concentration is depleted, is called a diverging node of the zonal wind. At altitudes above 120 km, where redistribution of the ions occurs owing to the action of the meridional wind component, the converging node is located between the northward wind (above) and the southward wind (below), and the diverging node is located between the inversely directed winds. Let us denote the converging nodes as CM and CZ and the diverging nodes as DM and DZ for the meridional and zonal winds, respectively.

Long-duration observations of the winds in the meteor zone show that the semidiurnal harmonic provides the principal contribution to the mid- and high latitude tidal motions [Lysenko et al., 1994]. According to classic tidal theory the semidiurnal tide is a superposition of various modes [Chapman and Lindzen, 1972; Hines, 1995], the main contribution to the semidiurnal tide at middle latitudes being provided by the (2, 2), (2, 4) and (2, 6) modes [ Fakhrutdinova and Ishmuratov, 1991]. In the northern hemisphere, the spatial hodograph of the wind velocity of any mode is a left-handed corkscrew [ Akchurin et al., 1995]. Thus the nodes in the tidal mode for one oscillation period are situated along the vertical in the following order (up from below): DM, CZ, CM, and DZ (any cyclic permutation of the nodes is possible if the mode is a propagating wave). Each tidal mode has its own value of downward phase velocity (the velocity of any node of the mode), in particular: ~12.5 km h-1, ~4.0 km h-1, and ~2.5 km h-1 for the (2, 2), (2, 4), and (2, 6) modes, respectively. When it is necessary in the following arguments to emphasize that the chosen node belongs to a particular mode of the semidiurnal tide, the mode number in parentheses will be added to the name of the node. For example, DZ(2) and DZ(6) denote the diverging nodes of the (2, 2) and (2, 6) zonal modes, respectively.

If there is only one tidal mode, WST predicts that the Es layers should be formed and descend in the converging nodes of this mode. Initially, the Es layer would move together with the CM node down to 120-130 km altitude, and then it would move together with the CZ node down to 90-100 km altitude. At altitudes of about 120 km (where the meridional convergence is changed by the zonal one) an accelerated motion of the Es layer from the CM node to the CZ node with a velocity which exceeds the phase velocity of the given mode, should be observed [Chimonas, 1973].

Let us call the Es descents with velocities which exceed the phase velocity of the (2, 2) mode, "superfast" descents; the descents with equal velocities the "fast" descents, and the descents with lower velocities the "slow" descents.

The descent of the Es layer will occur down to a definite altitude, below which the converging CZ node would not be able to entrain this layer because of the increase in the collision frequency of the ions. This is called the damping altitude and depends on the phase velocity of the tidal mode and the atomic mass of the transported ions [ Chimonas and Axford, 1968].

For tidal modes with phase velocities above 3.6 km h-1 the damping height is about 100 km, and for modes with lower velocities the damping height is 90-95 km [ Mathews and Bekeny, 1979]. Actually there exist in the atmosphere not only various tidal modes, but also gravity waves of nontidal origin. Thus the behavior of the Es layers could well be different from that described above.

Data from the Obninsk meteor radar (which does not measure echo altitude) provide the mean wind velocity over the meteor zone at an average altitude of 95 km [ Lysenko et al., 1994]. A determination of the vertical position of the node points is possible only by using the above theoretical representation of the semidiurnal tide.

4. Results of Initial Data Processing

Link to Fig. 1 An example of typical diurnal variations of the real height and frequency parameters of the Es layers is presented in Figure 1 as observed on July 3, 1993 (the afternoon), and July 4, 1993 (the entire day). To reveal the possible influence of various modes of the semidiurnal tide on variations in the Es layer altitude, Figure 1 shows the trajectories of descent of various nodes of the (2, 2) (dashed lines) and (2, 6) (dashed-dotted lines) modes. These trajectories are drawn in such a way that they intersect the 95-km altitude when the corresponding node (the name of the node is shown in the upper part of the trajectory) was measured by the meteor radar. To simplify the figure, the descent of the nodes of the (2, 6) mode is represented only by the descents of the CZ node, and the descents of the nodes of the (2, 4) mode are not shown at all. Most of the time the Es layers of the "main" type (thin line in Figure 1a) descend with a velocity close to the phase velocity of the (2, 6) mode. However, the Es layers of the "accompanying" type (the thick line) often shift downward with a velocity either equal to or exceeding the phase velocity of the (2, 2) mode. Part of these "fast" or "superfast" descents of the "accompanying" Es (marked by the double line under the horizontal axis in Figure 1a) occur at approximately the same altitudes and at the same time as the descents of the DM(2) node (at 0500-0600 LT in the morning and at 1800-1900 LT in the evening). This fast downward motion of the Es layer cannot be accounted for by a transition of the layer from the CM(2) node to the CZ(2) node - the radar measurements of nodal times do not match. Nor can the motion be explained by the action of the DM(2) node as described by the WST approximation. The DM node is divergent, so formation and motion of the Es layer with this node are impossible. However, the accumulation of positive ions near (above) the DM node, where the maximum of the eastward wind is located, is possible owing to the change with height of the effect of the Earth's magnetic field on ion motion. For example, accumulation of the ions at some altitude (above 120 km) under the action of only the homogeneous eastward wind would occur owing to the fact that the ions from the lower layers are accelerated to higher vertical velocity than the ions from the upper layers because the magnetic field effect increases with altitude. The divergence of the positive ion concentration above the DM(2) node is weakened by the two processes described above, in which two components of the zonal thermospheric wind participate: the prevailing wind (via the vertical gradient of the magnetic field effect) and the semidiurnal component (via the vertical gradient of the zonal wind velocity favorable for ion compression). The efficiency of these processes, which compensate for ion divergence, grows as the DM(2) node approaches the altitude of 120 km. So the point of total compensation of the divergence would move down with a velocity exceeding the phase velocity of the tide. The trajectory of the motion of the layers formed at this point would be steeper than the trajectory of the DM node (see 0500-0600 LT on July 4 in Figure 1). The competition between the actions of the meridional and zonal components of the thermospheric wind leads to the formation of low intensity layers. Figure 1 demonstrates that the Es layers formed by the action of the DM node (at 0500-0600 LT or 1800-1900 LT) are located above or near this node (Figure 1) and have small critical frequency values (in Figure 1 they exceed foE by only 0.2-0.5 MHz).

The trajectories of other "fast" descents of the Es layers are due to the CZ(2) nodes in the 100-120 km interval (shown by the single line below the horizontal axis in Figure 1). Such descents occur at 1000-1200 LT and 2100-2400 LT (except for the "superfast" descent of the "accompanying" layer at 2100-2200 LT on July 3). Figure 1 shows that the Es layers are able to move either under the action of only the CZ(2) node or under the action of two nodes (CZ(2) and CZ(6)) with later merging of the Es layers, which existed at these nodes. The damping altitude for the (2, 2) mode is higher than that for the (2, 6) one; so the Es layer formed after the merging of the CZ(2) and CZ(6) nodes moves further with the CZ(6) node.

The limiting frequencies of the "main" and "accompanying" Es layers exceed the foE values by about 2-3 times when the CZ(2) node passes the 100-120 km height interval (Figure 1). When the CZ(2) node descends to 95 km (these times are indicated in Figure 1 by the vertical dashed lines), the values of the limiting frequencies rapidly decrease and become equal to foE.

The trajectories of the descents of the CZ(6) nodes (shown by dashed-dotted lines) in Figure 1 are theoretical, because the absence of an altimeter at the Obninsk meteor radar installation does not allow reliable determination of the phase of this mode. This fact may explain the small discrepancy in time (about 2 hours) between the trajectories of the "slow" descents of the CZ(6) node and the Es layer (Figure 1).

Thus the short interval between the vertical sounding sessions (15 min) and detailed analysis of the ionograms has made it possible to reveal the action of the (2, 2) mode on the Es layer motions, which has not been done in previous analyses of the hourly data of the stations of the ionospheric network [ Fomichev and Shved, 1981; Gorbunova and Shved, 1984]. This tidal mode is often observed in summer in measurements of the wind in the meteor zone [ Lysenko et al., 1994].

Unfortunately, the processes in the lower thermosphere are significantly influenced not only by the wind shears, but also by various photochemical processes, particle precipitation, intensification of the external electric field, nonlinear interaction of the internal gravity waves (IGW), etc. That is why a clear relationship between variations of the ionospheric and wind parameters as described above is not observed continuously. To study this relationship a statistical analysis of the monthly set of data on frequency and height parameters of the Es layer has been carried out.

Link to Fig. 2 The hourly mean values of the vEs velocity of descent of the Es layers and the probability of their appearance P(vEs) for the velocity intervals >15 km h-1, 10-15 km h-1, 4-10 km h-1, and 0-4 km h-1 were calculated to provide a more accurate determination of the influence of the modal composition of the semidiurnal tide on Es layer motions. The results of the calculations for the "main" and the "accompanying" Es layers are presented in Figures 2a and 2b, respectively. The limits of the above intervals were chosen in such a way that curve 1 (v > 15 km h-1) characterizes the probability of the Es layer descending with a velocity which exceeds the phase velocity of the (2, 2) mode, and curves 2, 3, and 4 (v < 15 km h-1) characterize descents with velocity close to the phase velocities of the (2, 2), (2, 4), and (2, 6) modes, respectively.

Diurnal variations of the frequency parameters of the Es layers were studied on the basis of the parameter dfoEs = foEs - foE, where foE is the critical frequency of the E layer. This parameter makes it possible to exclude the influence of the background ionization in the diurnal variations of foEs. Figures 2c and 2d show the probability of appearance P(dfoEs) for dfoEs exceeding the values 5 MHz, 3 MHz, 1 MHz, 0.3 MHz, and 0.1 MHz for the "main" and "accompanying" layers, respectively.

Let us consider the diurnal variations of the frequency and height parameters of the "main" Es layer. Analysis of the set of P(vEs) curves (see Figure 2a) shows that the most probable descent velocity of the Es layers is at a velocity close to the phase velocity of the (2, 6) mode. The probability of motion with this velocity (curve 4) is equal to about 45%, a value that exceeds by 2-3 times the probability of downward motion with velocities close to the phase velocities of the (2, 2) and (2, 4) modes. Therefore the "main" Es layer is apparently located in the CZ(6) converging node.

However, the diurnal variations of the frequency parameters indicate a significant influence of the DZ(2) and CZ(2) nodes of the zonal wind. For example, at 0900-1100 LT and 2100-2300 LT in the periods of CZ(2) node passage of the region of zonal convergence (marked by the single line under the horizontal axis in Figure 2c), an increase of the Es layer intensity occurs. Times of minima in the curves of the P(dfoEs) family at 0600 LT and 1800 LT (the double line under the horizontal axis in Figure 2c) can be related to a passage of the 95-100 km interval by the DZ(2) node. Disappearance of the Es layers from the 95-110 km interval due to this node makes it possible to observe in the ionograms descents of the "accompanying" Es layers due to action of the wind system near the DM(2) node (such events are marked by the double line under Figure 1a).

As was mentioned previously, the "accompanying" Es layer consists of Es layers of several types (for example, consequent and additional). That is why the set of the P(vEs) curves for the "accompanying" Es layer (Figure 2b) demonstrates more complicated diurnal variations than that of the "main" Es layer. Using the results of the analysis of the diurnal variations presented above, one can state that the maximum of curve 1 (superfast descent) at 0600 LT is due to the action of the diverging DM (the double line under the horizontal axis in Figure 2d). The evening (superfast) descent of this node at 1800 LT is not manifested in curve 1 as in the morning because it is screened by the Es layers located below during the beginning of the descent. A typical value of dfoEs for the "accompanying" Es layer formed by the DM(2) node is 0.3-0.5 MHz. The probability of appearance of layers with such values of dfoEs is characterized by the distance between curves 4 and 5 (Figure 2d). The distance maximizes at the same times (0600 LT and 1800-1900 LT).

Though action of the CZ(2) node (marked by the single line under the horizontal axis in Figure 2d) is clearly manifested in variations of curves 1-3 of the P(vEs) family for the "accompanying" Es layer in the evening, in the daytime the action is more weakly manifested in curve 2 (~5%) owing to screening by the layers located below. Analysis of the evening maxima in curves 1-3 (at 2200-2300 in Figure 2b) shows that the probability of superfast descents exceeds the probability of fast ones by a factor of 2. This fact may be explained by superfast transitions of the Es layers between the CM(2) and CZ(2) nodes, which are observed at night owing to the absence of screening by the layers located below. However, action of the CZ(2) node in the diurnal variations of curves of the P(dfoEs) family for the "accompanying" Es is manifested both in the daytime and in the evening (at 0900-1100 LT and 2200-2300 LT in Figure 2d). The other maxima (at 0200 LT and 1500 LT) in the curves of the P(vEs) and P(dfoEs) families for the "accompanying" Es layers ( Figures 2b and 2d) apparently are related to passage of the CM(2) node. The CM(2) node is convergent above 120 km; the Es layers formed in it should be destroyed at altitudes below 115 km. This can be seen in Figure 1a at 0300 LT on July 4 and at 0200 LT on July 5 (the events are marked by the asterisks under the horizontal axis).

Upward motions of the Es layers of all types (not shown in Figure 1) in most cases are apparent. They may be caused by appearance of lateral reflections, or by reflections from the layers above and interruption (weakening) of reflections from the layers situated below or closer. However, as was mentioned by Mathews and Bekeny [1979], the upward motions may also be related to the destruction in some particular altitude range of some tidal mode and transportation of the long-lived ions from its converging node into a node of another mode located above.

Thus analysis of the curves of the P(vEs) and P(dfoEs) families of the "main" and "accompanying" Es layers confirms the conclusions regarding the influence of the (2, 2) mode nodes on Es layer formation, which have been made earlier on the basis of consideration of the diurnal variations of the real height and critical frequency of the Es layer for particular days.

5. Discussion

Figure 2 shows that the action of the semidiurnal tide (especially the (2, 2) mode) is permanently manifested in variations of the Es layer parameters and is apparently due to a very long vertical wavelength (150 km). To form an intense Es layer with a typical thickness of about 3-5 km, high values of wind shear in the converging node of the tidal mode are needed, and should correspond to rather high amplitudes of the tidal mode velocity. It is more realistic to suggest that formation of such a layer would happen owing to either the small-scale tidal modes or the IGW of a nontidal nature. In that case the (2, 2) semidiurnal mode of the tide first makes some sort of "collection" of the long-lived ions, which are needed to form an intense layer, and accumulates them in the region of its converging node. Then more small-scale IGW with a high value of wind shear, generated in this region, may continue compression of the ions accumulated by the tide into a narrow layer.

In addition to the above mechanisms, the following processes may influence formation and motion of Es layers: a sharp intensification of the external electric field [ Gershman et al., 1976], which is equivalent to an increase of the wind velocity in a converging node; interaction (nonlinear) of several IGW, which leads to a convergence of ions not only in the vertical direction, but in the horizontal direction as well [ Gershman et al., 1976]; and precipitation of energetic particles [ Chavdarov et al., 1975].

6. Conclusion

To explain the observed diurnal variations of parameters of the midlatitude Es layer in summer, one needs the presence of several tidal modes in the height interval 95-120 km. The diurnal variation of the altitude of Es layers corresponds to the motion of the converging node of the (2, 6) mode.

Many observed properties of the summer diurnal variations of the parameters of the midlatitude Es layer are explained by action of the (2, 2) semidiurnal mode, in particular: appearance at 0600 LT and 1900 LT of "extra fast" descents of the consequent Es layers in the 120-160 km height interval is due to action of the wind system near the DM node; appearance at 0900-1100 LT and 2100-2300 LT of "fast" descents of either the "main" Es layer (if the "accompanying" one is absent) or the "accompanying" one with consequent merging with the "main" layer at altitudes of 100-120 km is due to action of the CZ node; the values of foEs for these Es layers are high at 0000 LT and 0900-1200 LT, when the CZ node passes the 100-115 km altitude interval, and decreases to foE, when the CZ node descends to an altitude of 95 km.


This work was supported by the Russian Foundation for Basic Research (project 94-05-16090a).


Akchurin, A. D., et al., The influence of lower thermosphere dynamics on an appearance of the sporadic E layer, Geomagn. Aeron., 35 (2), 123, 1995.

Chapman, S., and R. S. Lindzen, Atmospheric Tides, D. Reidel, Dordrecht, 1970.

Chavdarov, S. S., Yu. K. Chasovitin, S. P. Chernysheva, and V. M. Sheftel, The Midlatitude Sporadic Layer of the Ionospheric E Region, 120 pp., Nauka, Moscow, 1975.

Chimonas, G., Wind component exchange and the rapid vertical movement of a sporadic E layer, J. Geophys. Res., 78 (25), 5636, 1973.

Chimonas, G., and W. I. Axford, Vertical movement of temperate-zone sporadic E layers, J. Geophys. Res., 73 (1), 111, 1968.

Fakhrutdinova, A. N., and R. A. Ishmuratov, Seasonal variations of the contribution of modes in the semidiurnal tidal motions in the midlatitude lower thermosphere, Geomagn. Aeron., 31 (5), 904, 1991.

Fomichev, V. I., and G. M. Shved, On the height and probability of formation of the Es layer at moderate latitudes, Geomagn. Aeron., 21 (4), 630, 1981.

Gershman, B. N., Yu. A. Ignat'ev, and G. Kh. Kamenetskaya, Mechanisms of Formation of the Ionospheric Sporadic E Layer at Various Latitudes, 108 pp., Nauka, Moscow, 1976.

Gorbunova, T. A., and G. M. Shved, Analysis of the semitransparence of the Es layer as an indicator of the turbulence under dynamically homogeneous conditions, Geomagn. Aeron., 24 (1), 30, 1984.

Hines, C. O., Atmospheric gravity waves, in Thermospheric Circulation, p. 85, Mir, Moscow, 1975.

Lysenko, I. A., et al., Wind regime at 80-110 km at mid-latitudes of the northern hemisphere, J. Atmos. Terr. Phys., 56 (1), 31, 1994.

Mal'tseva, O. A., S. P. Chernysheva, and V. M. Sheftel, Comparison of the real and virtual height of the Es sporadic layer, Geomagn. Aeron., 13 (2), 361, 1973.

Mathews, J. D., and F. S. Bekeny, Upper atmosphere tides and the vertical motion of ionospheric sporadic layers at Arecibo, J. Geophys. Res., 84\,(A6), 2743, 1979.

Minullin, R. G., et al., A Digital Ionospheric Facility "Tsyklon-9", Kazan' University Press, Kazan', Russia, 1994.

Portnyagin, Yu. I., Large-scale irregularities in the wind field at altitudes of 80-100 km, Izv. Acad. Sci. USSR Atmos. Okeanic Phys., 17 (3), 236, 1981.

Robinson, B. J., On some disturbances in the E region, J. Atmos. Terr. Phys., 19, 160, 1960.

Whitehead, J. D., Recent work on mid-latitude and equatorial sporadic Es, J. Atmos. Terr. Phys., 51 (5), 401, 1989.

 Load file for printing and local use.

Go to the top