- Email: [email protected]
Ionospheric vertical drift response at a mid-latitude station
Accepted Manuscript Ionospheric vertical drift response at a mid-latitude station Daniel Kouba, Petra Koucká Kní žová PII: DOI: Reference:
S0273-1177(16)30153-3 http://dx.doi.org/10.1016/j.asr.2016.04.018 JASR 12703
To appear in:
Advances in Space Research
Please cite this article as: Kouba, D., Kní žová, P.K., Ionospheric vertical drift response at a mid-latitude station, Advances in Space Research (2016), doi: http://dx.doi.org/10.1016/j.asr.2016.04.018
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Ionospheric vertical drift response at a mid-latitude station Daniel Kouba Institute of Atmospheric Physics ASCR, Bocni II 1401, 141 31 Prague, Czech Republic
Petra Kouck´a Kn´ıˇzov´a Institute of Atmospheric Physics ASCR, Bocni II 1401, 141 31 Prague, Czech Republic
Abstract Vertical plasma drift data measured at a mid-latitude ionospheric station Pruhonice ( 50.0◦ N, 14.6◦ E) were collected and analysed for the year 2006, a year of low solar and geomagnetic activity. Hence these data provide insight into the drift behaviour during quiet conditions. The following typical diurnal trend is evident: a significant decay to negative values (downward peak) at dawn; generally less pronounced downward peak at dusk hours. Magnitude of the downward drift varies during the year. Typically it reaches values about 20 m s−1 at dawn hours and 10 m s−1 at dusk hours. Maximum dawn magnitude of about 40 m s−1 has been detected in August. During daytime the vertical drifts increases from the initial small downward drifts to zero drift around noon and to small upward drifts in the afternoon. Night-time drift values display large variability around a near zero vertical drift average. There is a significant trend to larger downward drift values near dawn and a less pronounced decrease of the afternoon upward vertical drifts near sunset. Two regular downward peaks of the drift associated with the dawn and dusk are general characteristics of the analysed data throughout the year 2006. Their seasonal course corresponds to the seasonal course of the sunrise and sunset. The duration of prevailing negative drift velocities forming these peaks and thus the influence of the dawn/dusk on the drift velocity is mostly ∗
Daniel Kouba Email addresses: [email protected] (Daniel Kouba), [email protected] (Petra Kouck´ a Kn´ıˇzov´ a)
Preprint submitted to Advances in Space Research
April 27, 2016
1.5 - 3 hours. The dawn effect on vertical drift tends to be larger than the effect of the dusk. The observed magnitude of the sunrise and sunset peaks show significant annual course. The highest variability of the magnitude is seen during winter. High variability is detected till March equinox and again after September equinox. Around solstice, both peaks reaches lowest values. After that, the magnitudes of the drift velocity increase smoothly till maxima in summer (August). The vertical drift velocity course is smooth between June solstice and September equinox. In general, the detected values of the observed vertical drift are of lower magnitudes compare to low latitudes. Drift data in midlatitudes seems to be more influenced by the atmospheric waves than data in lower latitudes. Keywords: vertical plasma drift; Digisonde; Mid-latitude ionosphere; F-layer; daily pattern 1. Introduction Nowadays there are many ionospheric stations equipped with various ionosondes around the world. There are several models operating (Digisondes, IPS ionosondes, CADI, dynasondes), however practically Digisondes only measure the drifts on a regular base. A classical ionosonde uses the vertical ionospheric sounding to produce ionograms. The vertical ionospheric sounding provides the information about the profile of electron concentration. In general the ionospheric plasma is in motion, driven by electric fields and neutral winds. Information about the magnitude of these movements can be obtained using the direct drift measurements provided by modern digital ionosondes. The transmitted signal illuminates a large area of the ionosphere due to the antenna radiation pattern. Receiving antennas detect signal reflected from the ionosphere. For a perfectly smooth, vertically stratified ionosphere only one such location exists, responsible for a single vertical echo. However, large number of oblique echoes is usually detected. The non-vertical echoes occur due to ionospheric tilts near the solar terminator, modulations by atmospheric gravity waves, irregularities generated by plasma instabilities and other ionospheric phenomena. Detection of echoes from oblique directions is crucial for drift measurements. Spectral analysis is applied to the signal reflected from the ionosphere to distinguish individual echoes with different Doppler frequency shifts. For in2
dividual echoes heights of reflections, horizontal locations of reflection points in the ionosphere, values of Doppler shift, and signal amplitudes on receiving antennas are identified. Location of the reflection points can be graphically displayed in so called SKYmap (Reinisch et al., 2005). The technique for drift velocity determination is commonly referred as the ”Digisonde Drift Analysis” (DDA method) and the resulting velocity is called the ”drift velocity” (Reinisch et al., 1998). The DDA method is implemented in the software tool ”Drift Explorer” (Kozlov and Paznukhov, 2008) commonly used by Digisonde users. The automatic data processing software is distributed with the Digisondes for plotting the real-time SKYmaps and automatic drift velocity computing. Automatically scaled data are available through the Digisonde web pages and DriftBase (Reinisch and Galkin, 2011). For a given month Altadill et al. (2007) obtained daily pattern of the vertical velocity component by computing the monthly median values for mid-latitudinal station Ebro. Their study indicates daytime values usually close to zero with low variability. Night-time values display larger variability. The most distinct and systematic feature occurs after sunrise when vertical velocity decreases rapidly to a negative value of about -20 to -50 m/s. Belehaki et al. (2006) obtained the daily drift pattern for mid-latitudinal station Athens during quiet geomagnetic conditions. They detected a daily pattern in horizontal components of drift velocity vector. However the vertical component tends to present only a weak diurnal pattern with small amplitude. It is possible to determine the complete drift velocity vector using the described Digisonde drift measurements. When we focus on the vertical component of the drift velocity only, there are several other possibilities. Using multi-frequency HF Doppler radar observation is another possibility for direct measurement of the drift velocity vertical component (Prabhakaran Nayar et al., 2009; Mathew et al., 2010). The advantage of this technique in comparison with multifunctional Digisonde is usually the better time resolution. Mathew et al. (2010) deal with pre-sunrise and post-sunset characteristics of equatorial vertical plasma drift measured with multi-frequency HF Doppler radar. They observed the post-sunset enhancement and quasi-periodic fluctuations with periodicities 20-32 min. during post-sunset period. The plasma drift during pre-sunrise period shows enhancement prior to the ground sunrise followed by downward excursion. Incoherent scatter radar (ISR) is another instrument usable for direct mea3
surement of ionospheric drifts (Woodman and Hagfors, 1969). Jicamarca ISR can continuously observe vertical and zonal components of ion velocities. Based on the long term Jicamarca ISR measurements, a model of drift velocity was developed (Scherliess and Fejer, 1999). This is a widely used model for the equatorial region. During the day the upward direction is dominant and during night the downward direction of vertical drift in equatorial F region prevails in the quiet conditions (Fejer at al., 1991). There are also other models obtained using different measurements, the list can be found in Adeniyi et al. (2014). The height-resolution measurement of coherently scatter echoes occurring during daytime hours at around 150 km altitude (Kudeki and Fawcett, 1993) is the alternative to monitoring the equatorial vertical drifts by ISR. The experiment using ROCSAT-1 cannot be omitted in the list of the direct methods used to obtain information on vertical ionospheric drift (Huey Ching Yeh et al., 1999; Fejer et al., 2008). The indirect estimation based on the temporal evolution of the measured ionospheric characteristics is also often used for the calculation of the vertical drift component. The vertical velocity is estimated according to the change of characteristics, usually reflection frequency or corresponding electron concentration, scaled from the classical quarter-hour-ionograms. The cadence of ionogram measurements (typically 15 minutes) can be problematic. The drift velocity is calculated for each time interval between ionogram measurements. This assumes that the change of drift velocity will be slower than the time difference between the individual ionograms. It may be satisfying for quiet conditions but there are many detectable processes which are much faster volatile - geomagnetic storms, TID, etc. The advantage of this method is that it can be used for historical data or classical ionosondes which are not able to provide direct drift measurements. For example the value of (dh0 F/dt) is then called a vertical drift velocity (Batista et al., 1986; Saranya et al., 2014). This method is often limited to the measurements in the evening hours when the F-layer height is over 300 km (Batista et al., 1996; Subbarao et al., 1994). Bittencourt and Abdu (1981) found that for special time periods during sunset and evening hours, Fregion exceed a threshold height 300 km, the apparent vertical displacement velocity of the F-region, inferred from ionosonde measurements, correspond to vertical E × B plasma drift velocity determined by incoherent backscatter radar measurements in the F region. Below a threshold the recombination 4
processes affect significantly the plasma motion. Good agreement between Digisonde and incoherent scatter radar drift measurements during sunset and evening hours are reported by Bertoni et al. (2006). Additionally they found the correspondence during post-midnight hours (between 02 and 03 LT) and sunrise hours (until 08 LT). Adebesin et al. (2013) investigated seasonal behaviour of the vertical drift. They concluded, that vertical drifts obtained by Digisonde measurements only match the E × B drift if the Fregion is higher than 300 km is reliable, but does not hold for the night-time period of 22 − 06 LT under condition of solar minima. There are several variations of indirect methods using the temporal evolution of the ionogram characteristics. Instead of the h0 F some authors use for their studies parameter hmF 2 (Adeniyi et al., 2014; Adebesin et al., 2015). Another possibility is the measuring of the virtual height at fixed frequency from the F-layer trace on ionogram. Such measurement was used to compare the indirect fixed frequency results with the multi-frequency HF Doppler radar in Trivandrum. For the pre-sunrise period the results are in good agreement, the biggest difference occurs during sunrise(Prabhakaran Nayar et al., 2009). Based on the above mentioned methods it is evident that lots of authors study ionospheric vertical drifts using many different methods. Most of the research and articles are focused on the equatorial or low midlatitudes. A review of the measurements and theoretical models in low latitudes is provided in Fejer (1991). In midlatitudes the ionosphere monitoring is provided by Digisondes and ionosondes. Hence, knowledge of the vertical drifts in the mid-latitudes is still much more limited. 2. Data and methodology In January 2004 a new Digisonde DPS-4 with four cross-loop receiving antennas was installed in the observatory Pruhonice. The Digisonde DPS4 at Pruhonice provides routine vertical ionospheric sounding with a time resolution 15 minutes, except special campaigns of higher sampling rate. The critical frequency (foF2) is automatically detected by an ionogram autoscaling procedure ”Artist” (Galkin et al., 1996, 2008). Data and autoscaled characteristics are distributed to the data centres and presented in real-time on web page http://digisonda.ufa.cas.cz. In addition to the ionogram sounding the drift measurements are performed in both F and E regions. 5
For the F-region drift measurement it is the most convenient to perform an autodrift mode in a short order after ionogram sounding (Reinisch et al., 2005) because the sounding frequencies are automatically selected below the autoscaled foF2 frequency. In our schedule1 the F-region drift measurement follows 5 minutes after the ionogram sounding. The height resolution is set to 5 km and the Doppler shift spectral resolution is 0.195 Hz. Detail description of the measurement is provided in the work Kouba et al. (2008). The output of the measurement in the autoscale mode must be futher manually checked. An employment of the automatic output of the measurement may lead to mixing the reflections from different ionospheric layers during daytime (E and F layers) and during night-time during for instance occurrence of sporadic E layer. Hence, it is important to manually check all the drift data and control them together with the preceding ionogram. The reflection points from raw skymaps were chosen using the selection method described in Kouba et al. (2008). The routine use of the automatically computed velocities may lead to inappropriate smoothing of the effects and/or creation of the artificial effects. In this work we use the reflections from F-region only. The initial drift observations at Pruhonice in 2004 revealed that in the automatic mode measurement it is necessary to control data manually. The DDA method encounters problems especially when range of the measured Doppler frequency shifts is too wide, multiple-hop Es or F reflections are present on ionogram, and velocity distribution across the sounding range is non-uniform Kouba et al. (2008). The automatic drift data processing is significantly improved in the new version of Drift Explorer software (Kozlov and Paznukhov, 2008). However for detailed analysis and research purposes the manual data processing is still inevitable. In this study we analyse data measured in 2006 at Pruhonice station (50 N, 14.6 E). Data set (1 January - 31 December, 2006) represents measurements performed during a period of low solar activity. Maximum value of the solar flux f10.7 cm reached its maximum of 89 sfu (observed) in April. There occurred only 4 short events of high geomagnetic activity classified as a storm events with kP > 5. Such an exceptionally long period of low activity allows us to study the behaviour of the drift velocity during quiet conditions in mid-latitude sector. The number and spatial distribution of detected reflection points relates to 1
described schedule for 2006
6
the current state of the ionosphere. Typically, during the spread F situation the good quality drift data are obtained (Kouba and Kouck´a Kn´ıˇzov´a, 2012). Due to departures from plane stratification the conditions are favourable for registration of large amount of reflected signals. Hence, it allows the accurate velocity estimation. If the ionosphere is stratified into horizontal layers without any spatial disturbances (according to quality of ionogram), we do not detect high number of reflection points necessary for estimating the drift velocity vector. The quantity and the spatial distribution of the detected reflection points mainly affect the accuracy of the horizontal drift components determination. The detected reflection points tend to cumulate near the vertical direction when the ionosphere is horizontally stratified. Such a situation does not allow the estimation of horizontal components. However, when we focus on the vertical velocity component only there is no problem in precise determination. Therefore in this paper we focus on the vertical component of drift velocity vector only. Our database consist of 32 130 (out of 35 040 measurements) manually checked F-region drift velocity vectors. The rest of the measurements were excluded due to the presence of Es layer or ionogram autoscaling failures leading to the incorrect drift measurement. Such a large data set allows us to study the statistical behaviour of velocity vector in quiet conditions. In order to obtain characteristic behaviour of the drift data it is necessary to eliminate influence of gaps with longer duration and/or sudden irregular wave-like events. Except drift velocities directly computed from measurement we use the mean velocities computed with 9 days long window. With this approach we get rid of breaks in the diurnal course of the drift. 3. Results Basic characteristics of the measured vertical velocities are presented in a histogram of these data values (Fig.1). Positive vertical drifts refer to upward plasma motion, while negative vertical drifts refer to downward plasma motion.The preliminary results for the first half of 2006 were published and the quality of the measurements was discussed (Kouba and Kouck´a Kn´ıˇzov´a, 2012). The histogram is well represented by a normal probability density function characterized by the following parameters: median v˜z = −0.35 m s−1 , mean value µ = hvz i = −0.301 m s−1 , and standard deviation σvz = 9.44 m s−1 . 7
Figure 1: Vertical velocity component histogram for F region measurements during 2006.
Someone might infer from the histogram that the vertical drift velocity in mid-latitudinal ionosphere is zero during quiet conditions. Any fluctuation around zero velocity would be caused by irregularities - TID, acoustic-gravity waves, etc. However, the detailed investigation based on the average daily trends does not confirm this expectation.
Figure 2: One-day course of the vertical drift for 2006, Feb 24: a) measured velocities, b) measured velocities smoothed using 30 minutes averaging window, c) averaged velocity measured over 9 days
On the upper panel of Fig.2 there is an example of a one-day-course of the vertical drift component for 2006, Feb 24. Displayed values of the verti8
cal velocity component were calculated for each successfully performed drift measurement without any smoothing of the data. The data show high dispersion. Hence, there is no evident diurnal trend and further process is necessary to derive diurnal course. In general, velocity reaches lower values during the daytime in comparison with the nigh-time values. Tendency of the drift to grow from negative values around sunrise up to positive values close to sunset could be found in the plot. On the middle panel there are the measured velocities displayed as it is typical on the Digisonde web pages. The moving average filter with 30 minutes window is applied on the time series of drift values. The resulting time series is not a smooth line. There are significant breaks and points departuring from the general course. These irregularities occur when gaps, caused by drift measurement failure, are longer than the filter window duration. However, this figure shows a smaller velocities during daytime hours, a higher probability of small negative values in the morning and more positive values in the afternoon. It shows that higher velocity values and stronger wave activity are observed during the night time. Bottom panel of Fig.2 shows the diurnal course of the mean vertical velocity measured for 9 consecutive days. As a representative values for Feb 24 we consider mean values computed for four preceding and four subsequent days (Feb 20 - Feb 28). Comparing the middle and bottom panel we see that both plots are in qualitative agreement. On the bottom plot, there are no discontinuities. Hence the diurnal course is well seen from the plot. The most important feature is that the drift velocity significantly fall to negative values near dawn. The drift velocity reaches its minimum close to the local sunrise. Around the dusk time a similar decrease is observed forming secondary minima of the diurnal course shortly close to sunset. During the daytime, gradual increase of the vertical velocity from negative value after sunrise over the zero around noon to a positive value before sunset is well seen. Averaging groups of nine-days was performed for the data for whole year 2006. On Fig.3 the averaged one-day courses are displayed for the whole year 2006. The following features of vertical velocity diurnal cycles were found: • significant decline (downward peak) around dawn, • less pronounced downward peak around dusk, • increase from small negative values after sunrise over the zero around 9
Figure 3: Average daily trend of the vertical drift for 2006 - seasonal change of the vertical velocity component for 9 days averages
noon to the small positive values before dusk, • no regular course during the night; greater velocity values and higher wave activity. In the diurnal course two regular downward peaks were associated with the dawn and dusk. The existence of the peaks is a general characteristic of all analysed data throughout the year 2006. This fact is evident from Fig.3. Each day the minima of the vertical drift for dawn and dusk were identified on the diurnal course. Seasonal changes of occurrences of the time of these peaks demonstrates Fig.4. It corresponds to the seasonal course of the sunrise and sunset. On the plot, time of the peak minima is emphasised by larger symbols. The beginning of the sudden decrease and the return to normal values before the decrease are marked with smaller symbols. Seasonal course of peaks time occurrences shown on Fig.4 demonstrates linking with dawn and dusk. The peaks are closest to each other in the period around the winter solstice and the farthest apart around the summer solstice. The duration of prevailing negative drift velocities forming these peaks and thus the influence of the dawn/dusk on the drift velocity is mostly 1.5 - 3 hours. It should be noted that around the summer solstice (about May 20 - July 20) the sun does not set for F-region ionospheric heights at the latitude of the Observatory Pruhonice (50 N, 14.9 E). However, the ionization is less effective due to change of the zenith angle. The peaks related to dawn and dusk 10
Figure 4: Seasonal change of the dawn and dusk downward peaks for 9-days averaged daily trend. Large symbols - minima of the peak, small symbols - start of the sudden decrease, return to the values before peak
remain present of the diurnal course. Likely, there is not only a local effect of sunrise and sunset forming the negative velocity structures. The contributing phenomena associated with sunrise and sunset is a passage of the solar terminator. The moving border of the sunlit and and shadow regions forms gravity wave structures important in a global scale (for instance Chimonas and Hines (1970); Somsikov (1991, 1995) ). Gravity wave enhancements associated with motion of the Solar Terminator were reported for instance by ˇ Altadill et al. (2001), Galushko et al. (1998), Cheng et al. (1992), Sauli et al. (2006). Boˇska et al. (2003) found stronger gravity wave activity related to the Solar Terminator in the morning hours. It corresponds to the experimental observation of solar eclipses in the ionosphere. Different atmospheric activity is usually observed during initial and recovery phases of the eclipse ˇ (Jakowski et al., 2008; Sauli et al., 2007) and others) On the plot of mean daily trends Fig.3. velocity enhancements can be seen in the vicinity of the well pronounced minima. However, on the contrary to the well-pronounced negative velocity peaks, the preceding and the following velocity enhancements are difficult to locate. Night-time course of the velocity shows higher day-to-day variability. The post-sunset and presunrise enhancements are not as regular effects as the velocity downward peaks. On the night-time course, the oscillations with periods in the gravity wave domain can be detected/observed. Hence, we focus only on the regular 11
downward velocity peaks.
Figure 5: Seasonal change of magnitudes of decrease for the dawn and dusk peaks; dawn peak - square, dusk peak - triangle
The dawn effect on vertical drift tends to be larger than the effect of the dusk. The observed magnitude of the sunrise and sunset peaks show significant annual course. During first part of the year till solstice, higher variability is evident on both sunrise and sunset peaks. Around solstice, both peaks reaches lowest values. After that, the drift magnitudes of the drift velocity increase smoothly till maxima in summer (August). The vertical drift velocity course is smooth between June solstice and September equinox. The variability increases again after September equinox. The observed wave-like oscillation during winter months within the drift data we attribute to the planetary waves influencing the ionosphere. Most of the observed planetary waves in the ionosphere are propagating from lower atmosphere. Review of the meteorological/lower atmosphere influence on the ionosphere is discussed in details in Laˇstoviˇcka et al. (2006). Much stronger filtering by stratospheric winds in summer is responsible for much smaller penetration of the tropospheric planetary waves to the lower ionosphere heights in summer, which reflects in the observed much smaller summer time planetary wave activity in the ionosphere. This is demonstrated in Fig.5. The magnitudes of decrease for the dawn peaks are marked with squares and for the dusk peak with triangles. The magnitude of the velocity decrease was obtained as a difference of vertical 12
velocity value before the sudden drop and the value in the minimum of the downward peak. Another characteristic feature of all average daily trend of the vertical drift are small velocity values detected during the daytime. Between the dawn and dusk peaks are detected vertical velocities typically within the interval ±10m/s. After the dawn peak recovery there are detected small negative velocity values between −6 to −9 m/s in the most cases. Then the value gradually increases and reaches zero around the local noon. In the afternoon the increase continues till typical values about 6 − 10 m/s before the dusk drop. There were not found any seasonal changes of parameters characterizing this linear increase of vertical velocity during the daytime. We suppose that during day-time during quiet solar and geomagnetic conditions we observe dominant effect of the decrease/increase of the reflection height due to increasing/decreasing zenith angle. Vertical velocities detected during the night hours (between the dusk and the dawn peak) show higher variability compare to day time values. Velocities are generally positive and larger than during the daytime. During night-time the occurrence of wave-like oscillations is a regular characteristics. Typically, the observed waves belong to gravity wave domain as it is seen on the Fig.2. Wave activity is also larger but without any registered regularity. 4. Discussion and Conclusion The drift data of one whole geomagnetically quiet year were processed. This allows us to study the regular behaviour of the drifts in quiet conditions and provides us a base for subsequent studies of the irregular activities during disturbed conditions such as geomagnetic storms, solar eclipses etc. Seasonal course of both sunrise and sunset downward peaks confirm the dominant role of planetary waves in the ionosphere during winter season and around equinoxes. During summer months the course of the sunrise and sunset velocity peaks are smooth without apparent wave-like structures. Apparently, we observe competing mechanisms driving the plasma. We observe plasma motion forced by E×B. The configuration of the geomagnetic field plays important role. The geomagnetic field dip angle is 65.8◦ . The atmospheric waves are another mechanism that significantly influence the plasma in midlatitudes. Within the drift data wave-like structures are well seen.
13
Comparing results from low-latitude ionosphere, the vertical drift in midlatitude differs significantly. In general, the detected values of the observed vertical drift are of lower magnitudes compare to low latitudes. Within data we detect stronger influence of the atmospheric waves than in lower latitudes. On the diurnal course, there is no structure that corresponds to the pre-reversal enhancement as it is observed in low latitudes. The most clearly developed structure is the morning downward velocity peak close to sunrise. This structure occurs regularly and shows clear seasonal dependence. Similar but less pronounced structure, also downward, could be found close to sunset. Smooth increase from sunrise peak values through zero up to positive values preceding the sunset downward peak could be considered as a regular diurnal course. As the measured value of the plasma drift represent the resulting forcing it is difficult to answer what mechanisms is dominant for each moment. We suppose, that during winter season, the plasma motion is forced by the planetary waves. During night-time, the influence of gravity waves is much stronger than during day-time. The dynamics play a more important role in the ionospheric F-region during night-time than during daytime due to the lack of photochemical control at night. During sunrise and sunset downwardpeaks events, the effect of increasing/decreasing ionization contribute to the processes. It coincides with the passage of the Solar Terminator. Hence, the vertical component of the drift reflects the change of the F-layer that could be attributed at least two mechanism. The exact clarification of each contributing mechanism calls for further theoretical study and go beyond the focus of this observational study. Histogram on fig.1 shows that the measured vertical velocity is most frequently in the range of ±20 m/s. Values between 20 and 40 m/s are reported rarely and speeds above 40 m/s can be considered as extreme. It is consistent with the result obtained by Belehaki et al. (2006) where detected amplitude of vertical velocity component seldom exceeds 20 m/s. Similar analysis has been performed by Altadill et al. (2007) on the data measured in 2004 and 2005 in the observatory Ebro (40.8◦ N , 0.5◦ E) by mean of Digisonde 256. The Ebro and Pruhonice stations have one hour difference in local time. Both Ebro and Pruhonice stations belong to mid-latitudes. On the vertical component of the drift velocity gradual increase from negative values close to sunrise to positive values close to sunset has been found in Ebro data as well. This structure is, however, better developed on the diurnal course of the vertical component than in the monthly median courses (Figures 3. and 14
4. in Altadill et al. (2007)). The drop connected with the dawn as a main characteristic of the average daily trend of the vertical drift is well seen in Ebro data. The diurnal course of the vertical component and amplitudes of the detected velocities are in agreement in both stations. Our results are in agreement with published measurements in midlatitudes. We suppose that qualitatively same diurnal course of the drift would be found in all midlatitude stations. In the studied year 2006 there were no significant abnormal ionospheric event/activity ( http://wdc.kugi.kyoto-u.ac.jp/wdc/Sec3.html geomagnetic data, ftp://ftp.swpc.noaa.gov/pub/weekly/RecentIndices.txt solar data) Studied year 2006 was an exceptional year with respect to the solar, geomagnetic and ionospheric activity. There occurred only few isolated events of higher activity. Year 2006 represents last part of the 23rd Solar cycle that ended in 2008. Hence the results of this work may be further used as a reference for drift analyses during special events. This study may be the basis for the evaluation of various phenomena on the drift motion in mid-latitudinal ionosphere. 5. Acknowledgement This work has been supported by Grant Agency of the Czech Republic No. 15-24688S. 6. References References Adebesin, B. O., Adeniyi, J. O., Adimula, I. A., Oladipo, O. A., Olawepo, A. O., Reinisch, B. W., Comparative analysis of nocturnal vertical plasma drift velocities inferred from ground-based ionosonde measurement of hmF2 and h’F. J. of Atmosph. and Sol.-Ter. Phys. 122 (2015), pp. 97107. Adebesin, B.O., Adeniyi, J.O., Adimula, I.A., Reinisch, B.W., Low latitude nighttime ionospheric vertical E B drifts at African region. Advances in Space Research, 52 (2013), pp. 2226-2237 Adeniyi, J. O., Adebesin, B. O., Adimula, I. A., Oladipo, O. A., Olawepo, A. O., Ikubanni, S. O., Reinisch, B. W., Comparison between African 15
equatorial station ground-based inferred vertical E x B drift, Jicamarca direct measured drift, and IRI model. Advances in Space Research, 54(7) (2014), 1629-1641. doi:10.1016/j.asr.2014.06.0 Altadill, D., Sole, J.G., Apostolov, E.M., Vertical structure of a gravity wave like oscillation in the ionosphere generated by the solar eclipse of August 11, 1999. Journal of Geophysical Research 106 (2001), pp. 21419-21428 Altadill, D., Arrazola, D., Blanch, E. F-region vertical drift measurements at Ebro, Spain. Advances in Space Research 39 (2007), pp. 691-698. Batista, I.S., Abdu, M.A., Bittencourt, J.A., Equatorial F region vertical plasma drifts: seasonal and longitudinal asymmetries in American sector. J. Geophys. Res. 91, 12 (1986), pp. 12055-12064 Batista, I.S., de Medeiros, R.T., Abdu, M.A., de Souza, J.R., Bailey, G.J., de Paula, E.R., Equatorial ionospheric vertical plasma drift model over Brazilian region. J. Geophys. Res. 101, 10 (1996), pp. 10887-10892 Belehaki, A., Tsagouri, I., Michalaeras, G., Herekakis, T, F region drift observations from Athens Digisonde. Radio Sci. 41, RS6S35 (2006), doi:10.1029/2005RS003322 Bertoni, F., Batista, I.S., Abdu, M.A., Reinisch, B.W., Kherani, E.A., A comparison of ionospheric vertical drift velocities measured by Digisonde and Incoherent Scatter Radar at magnetic equator. J. of Atmosph. and Sol.-Ter. Phys. 68 (2006), pp. 669-678, doi:10.1016/j.jastp.2006.01.002 Bittencourt, J.A. and Abdu, M.A., A theoretical comparison between apparent and real vertical ionization drift velocities in the equatorial F region. Journal of Geophysical Research 86 (1981), pp. 2451-2454, doi:10.1029/JA086iA04p02451 ˇ Boˇska, J., Sauli, P., Altadill, D., Sol, J. G. Alberca, L. F., Diurnal variation of gravity wave type oscillations at midlatitudes in the ionospheric F region. Studia geophysica et geodaetica 47 (2003), pp. 579-586 Cheng, K., Huang, Y.N., Chen, S.W., Ionospheric effects of the solar eclipse of September 23, 1987, around the equatorial anomaly crest region. Journal of Geophysical Research 97 (1992), pp. 103-111
16
Chimonas G. and Hines C.O., Atmospheric gravity waves induced by a Solar eclipse. J. Geophys. Res., 75 (1970), pp. 875-875 Fejer, B. G., de Paula, E. R., Gonzales, S. A., Woodman, F. R., Average vertical and zonal F-region plasma drifts over Jicamarca, J. Geophys. Res. 96 (1991), pp. 901-906 Fejer., B.G., Low latitude electrodynamic plasma drifts: a review. J. of Atmosph. and Sol.-Ter. Phys. 53, 8 (1991), pp. 677-693 Fejer, B. G., Jensen, J. W., Su, S., Quiet time equatorial F-region vertical plasma drift model derived from ROCSAT-1 observation. J. Geophys. Res. 113 (2008), A05304 Galkin, I. A., Reinisch, B. W., Ososkov, G. A., Zaznobina, E. G., Neshyba, S. P., Feedback neural networks for ARTIST ionogram processing, Radio Sci., 31 (1996), pp. 1119-1129 Galkin, I. A., Khmyrov, G. M., Kozlov, A. V., Reinisch,B. W., Huang, X., Paznukhov,V. V., The ARTIST 5, Radio Sounding and Plasma Physics, AIP Proceedings (2008) Galushko V. G., Paznukhov, V. V., Yampolski, Y. M., Foster, J. C., Incoherent scatter radar observations of AGW/TID events generated by the moving solar terminator. Ann. Geophys. 16 (1998), pp. 821-827. Jakowski, N., Stankov, S. M., Wilken, V., Borries, C., Altadill, D., Chum, ˇ J., Bureˇsova, D., Boˇska, J., Sauli, P., Hruˇska, F., Ionospheric behavior over Europe during the solar eclipse of 3 October 2005. Journal of Atmospheric and Solar-Terrestrial Physics, 70, 6 (2008), pp. 836-853, doi10.1016/j.jastp.2007.02.016 ˇ Kouba, D., Boˇska, J., Galkin, I. A., Santolk, O., Sauli, P., Ionospheric drift measurements: Skymap points selection. Radio Science 43 (2008), RS1S90 Kouba, D., Kouck´a Kn´ıˇzov´a, P., Analysis of digisonde drift measurements quality. Journal of Atmospheric and Solar-Terrestrial Physics, 90-91 (2012), pp. 212-221 Kozlov, A. and Paznukhov, V. V., Digisonde drift analysis software, in Radio Sounding and Plasma Physics, AIP Conf. Proc., 974 (2008), pp. 167-174 17
Kudeki, E., Fawcett, C. D., High resolution observations of 150 km echoes at Jicamarca, Geophys. Res. Lett., 20 (1993), pp. 1987-1990 Laˇstoviˇcka, J., Akmaev, R. A., Beig, G., Bremer, J., Emmert, J. T., Global change in the upper atmosphere. Science, Vol.314, No. 5803 (2006), pp. 1253-1254 Mathew, T. J., Prabhakaran Nayar, S. R., Ravindran, S., Pant, T., Characteristics of the equatorial F-region vertical plasma drift observed during post-sunset and pre-sunrise hours. Advances in Space Research, 46 (2010), pp. 618-625 Prabhakaran Nayar, S. R., Mathew, T. J., Sreehari, C. V., Sumod, S. G., Devasia, et al., Electrodynamics of the equatorial F-region ionosphere during pre-sunrise period. Ann. Geophys., 27 (2009), pp. 107-111 Reinisch, B. W., Scali, J. L., Haines, D. M. Ionospheric drift measurements with ionosondes. Annales Geophysicae 41(5-6) (1998), pp. 695-702 Reinisch, B. W., Huang, X. Galkin, I. A., Paznukhov, V. V., Kozlov, A. V., Recent advances in real-time analysis of ionograms and ionospheric drift measurements with digisondes. Journal of Atmospheric and SolarTerrestrial Physics 67(12) (2005), pp. 1054-1062 Reinisch, B. W., Galkin, I. A., Global Ionospheric Radio Observatory (GIRO), Earth Planets Space, 63 (2011), 377-381 Saranya, P. L., Prasad, D. S. V. V. D., Rama Rao, P. V. S., Ionospheric vertical drifts over an Indian low latitude station and its comparison with IRI-2007 vertical drift model. Advances in Space Research, 54, 6 (2014), pp. 946-954 ˇ Sauli, P., Abry, P., Altadill, D., Boˇska, J., Detection of the wave-like structures in the F-region electron density: Two station measurements. Studia Geophysica et Geodetica, 50, 1 (2006), pp. 131-146, doi:10.1007/s11200006-0007-y ˇ Sauli, P., Roux, S. G., Abry, P., Boˇska, J., Acoustic-gravity waves during solar eclipses: Detection and characterization using wavelet transforms. Journal of Atmospheric and Solar-Terrestrial Physics, 69, 17-18 (2007), pp. 2465-2484, doi:10.1016/j.jastp.2007.06.012 18
Scherliess, L., Fejer, B. G., Radar and satellite global equatorial F region vertical drift model. J. Geophys. Res., 104(A4) (1999), pp. 6829-6842 Somsikov, V. M., Waves in atmosphere excited by Solar Terminator. Geomagnetizm i Aeronomiya, 31 (1991), pp. 1-12 (in Russian) Somsikov, V.M., On the mechanism for the formation of atmospheric irregularities in the solar terminator region. J. Atmos. Terr. Phys., 57 (1995), pp. 75-83 Subbarao, K. S. V., Krishna Murthy, B. V., Post-sunset F region vertical velocity variation at magnetic equator. J. Atmos. Sol. Terr. Phys. 56 (1994), pp. 59-65 Woodman, R. F., Hagfors, T., Method for the measurement of vertical ionospheric motion near the magnetic equator by incoherent scattering, J. Geophys. Res., 75(5) (1969), pp. 1205-1212 Yeh, H. C., Su, S., Heelis, R. A., Wu, J., The ROCSAT-1 IPEI Preliminary Results: Vertical Ion Drift Statistics. TAO, 10 (4) (1999), pp. 805-820
19
S0273-1177(16)30153-3 http://dx.doi.org/10.1016/j.asr.2016.04.018 JASR 12703
To appear in:
Advances in Space Research
Please cite this article as: Kouba, D., Kní žová, P.K., Ionospheric vertical drift response at a mid-latitude station, Advances in Space Research (2016), doi: http://dx.doi.org/10.1016/j.asr.2016.04.018
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Ionospheric vertical drift response at a mid-latitude station Daniel Kouba Institute of Atmospheric Physics ASCR, Bocni II 1401, 141 31 Prague, Czech Republic
Petra Kouck´a Kn´ıˇzov´a Institute of Atmospheric Physics ASCR, Bocni II 1401, 141 31 Prague, Czech Republic
Abstract Vertical plasma drift data measured at a mid-latitude ionospheric station Pruhonice ( 50.0◦ N, 14.6◦ E) were collected and analysed for the year 2006, a year of low solar and geomagnetic activity. Hence these data provide insight into the drift behaviour during quiet conditions. The following typical diurnal trend is evident: a significant decay to negative values (downward peak) at dawn; generally less pronounced downward peak at dusk hours. Magnitude of the downward drift varies during the year. Typically it reaches values about 20 m s−1 at dawn hours and 10 m s−1 at dusk hours. Maximum dawn magnitude of about 40 m s−1 has been detected in August. During daytime the vertical drifts increases from the initial small downward drifts to zero drift around noon and to small upward drifts in the afternoon. Night-time drift values display large variability around a near zero vertical drift average. There is a significant trend to larger downward drift values near dawn and a less pronounced decrease of the afternoon upward vertical drifts near sunset. Two regular downward peaks of the drift associated with the dawn and dusk are general characteristics of the analysed data throughout the year 2006. Their seasonal course corresponds to the seasonal course of the sunrise and sunset. The duration of prevailing negative drift velocities forming these peaks and thus the influence of the dawn/dusk on the drift velocity is mostly ∗
Daniel Kouba Email addresses: [email protected] (Daniel Kouba), [email protected] (Petra Kouck´ a Kn´ıˇzov´ a)
Preprint submitted to Advances in Space Research
April 27, 2016
1.5 - 3 hours. The dawn effect on vertical drift tends to be larger than the effect of the dusk. The observed magnitude of the sunrise and sunset peaks show significant annual course. The highest variability of the magnitude is seen during winter. High variability is detected till March equinox and again after September equinox. Around solstice, both peaks reaches lowest values. After that, the magnitudes of the drift velocity increase smoothly till maxima in summer (August). The vertical drift velocity course is smooth between June solstice and September equinox. In general, the detected values of the observed vertical drift are of lower magnitudes compare to low latitudes. Drift data in midlatitudes seems to be more influenced by the atmospheric waves than data in lower latitudes. Keywords: vertical plasma drift; Digisonde; Mid-latitude ionosphere; F-layer; daily pattern 1. Introduction Nowadays there are many ionospheric stations equipped with various ionosondes around the world. There are several models operating (Digisondes, IPS ionosondes, CADI, dynasondes), however practically Digisondes only measure the drifts on a regular base. A classical ionosonde uses the vertical ionospheric sounding to produce ionograms. The vertical ionospheric sounding provides the information about the profile of electron concentration. In general the ionospheric plasma is in motion, driven by electric fields and neutral winds. Information about the magnitude of these movements can be obtained using the direct drift measurements provided by modern digital ionosondes. The transmitted signal illuminates a large area of the ionosphere due to the antenna radiation pattern. Receiving antennas detect signal reflected from the ionosphere. For a perfectly smooth, vertically stratified ionosphere only one such location exists, responsible for a single vertical echo. However, large number of oblique echoes is usually detected. The non-vertical echoes occur due to ionospheric tilts near the solar terminator, modulations by atmospheric gravity waves, irregularities generated by plasma instabilities and other ionospheric phenomena. Detection of echoes from oblique directions is crucial for drift measurements. Spectral analysis is applied to the signal reflected from the ionosphere to distinguish individual echoes with different Doppler frequency shifts. For in2
dividual echoes heights of reflections, horizontal locations of reflection points in the ionosphere, values of Doppler shift, and signal amplitudes on receiving antennas are identified. Location of the reflection points can be graphically displayed in so called SKYmap (Reinisch et al., 2005). The technique for drift velocity determination is commonly referred as the ”Digisonde Drift Analysis” (DDA method) and the resulting velocity is called the ”drift velocity” (Reinisch et al., 1998). The DDA method is implemented in the software tool ”Drift Explorer” (Kozlov and Paznukhov, 2008) commonly used by Digisonde users. The automatic data processing software is distributed with the Digisondes for plotting the real-time SKYmaps and automatic drift velocity computing. Automatically scaled data are available through the Digisonde web pages and DriftBase (Reinisch and Galkin, 2011). For a given month Altadill et al. (2007) obtained daily pattern of the vertical velocity component by computing the monthly median values for mid-latitudinal station Ebro. Their study indicates daytime values usually close to zero with low variability. Night-time values display larger variability. The most distinct and systematic feature occurs after sunrise when vertical velocity decreases rapidly to a negative value of about -20 to -50 m/s. Belehaki et al. (2006) obtained the daily drift pattern for mid-latitudinal station Athens during quiet geomagnetic conditions. They detected a daily pattern in horizontal components of drift velocity vector. However the vertical component tends to present only a weak diurnal pattern with small amplitude. It is possible to determine the complete drift velocity vector using the described Digisonde drift measurements. When we focus on the vertical component of the drift velocity only, there are several other possibilities. Using multi-frequency HF Doppler radar observation is another possibility for direct measurement of the drift velocity vertical component (Prabhakaran Nayar et al., 2009; Mathew et al., 2010). The advantage of this technique in comparison with multifunctional Digisonde is usually the better time resolution. Mathew et al. (2010) deal with pre-sunrise and post-sunset characteristics of equatorial vertical plasma drift measured with multi-frequency HF Doppler radar. They observed the post-sunset enhancement and quasi-periodic fluctuations with periodicities 20-32 min. during post-sunset period. The plasma drift during pre-sunrise period shows enhancement prior to the ground sunrise followed by downward excursion. Incoherent scatter radar (ISR) is another instrument usable for direct mea3
surement of ionospheric drifts (Woodman and Hagfors, 1969). Jicamarca ISR can continuously observe vertical and zonal components of ion velocities. Based on the long term Jicamarca ISR measurements, a model of drift velocity was developed (Scherliess and Fejer, 1999). This is a widely used model for the equatorial region. During the day the upward direction is dominant and during night the downward direction of vertical drift in equatorial F region prevails in the quiet conditions (Fejer at al., 1991). There are also other models obtained using different measurements, the list can be found in Adeniyi et al. (2014). The height-resolution measurement of coherently scatter echoes occurring during daytime hours at around 150 km altitude (Kudeki and Fawcett, 1993) is the alternative to monitoring the equatorial vertical drifts by ISR. The experiment using ROCSAT-1 cannot be omitted in the list of the direct methods used to obtain information on vertical ionospheric drift (Huey Ching Yeh et al., 1999; Fejer et al., 2008). The indirect estimation based on the temporal evolution of the measured ionospheric characteristics is also often used for the calculation of the vertical drift component. The vertical velocity is estimated according to the change of characteristics, usually reflection frequency or corresponding electron concentration, scaled from the classical quarter-hour-ionograms. The cadence of ionogram measurements (typically 15 minutes) can be problematic. The drift velocity is calculated for each time interval between ionogram measurements. This assumes that the change of drift velocity will be slower than the time difference between the individual ionograms. It may be satisfying for quiet conditions but there are many detectable processes which are much faster volatile - geomagnetic storms, TID, etc. The advantage of this method is that it can be used for historical data or classical ionosondes which are not able to provide direct drift measurements. For example the value of (dh0 F/dt) is then called a vertical drift velocity (Batista et al., 1986; Saranya et al., 2014). This method is often limited to the measurements in the evening hours when the F-layer height is over 300 km (Batista et al., 1996; Subbarao et al., 1994). Bittencourt and Abdu (1981) found that for special time periods during sunset and evening hours, Fregion exceed a threshold height 300 km, the apparent vertical displacement velocity of the F-region, inferred from ionosonde measurements, correspond to vertical E × B plasma drift velocity determined by incoherent backscatter radar measurements in the F region. Below a threshold the recombination 4
processes affect significantly the plasma motion. Good agreement between Digisonde and incoherent scatter radar drift measurements during sunset and evening hours are reported by Bertoni et al. (2006). Additionally they found the correspondence during post-midnight hours (between 02 and 03 LT) and sunrise hours (until 08 LT). Adebesin et al. (2013) investigated seasonal behaviour of the vertical drift. They concluded, that vertical drifts obtained by Digisonde measurements only match the E × B drift if the Fregion is higher than 300 km is reliable, but does not hold for the night-time period of 22 − 06 LT under condition of solar minima. There are several variations of indirect methods using the temporal evolution of the ionogram characteristics. Instead of the h0 F some authors use for their studies parameter hmF 2 (Adeniyi et al., 2014; Adebesin et al., 2015). Another possibility is the measuring of the virtual height at fixed frequency from the F-layer trace on ionogram. Such measurement was used to compare the indirect fixed frequency results with the multi-frequency HF Doppler radar in Trivandrum. For the pre-sunrise period the results are in good agreement, the biggest difference occurs during sunrise(Prabhakaran Nayar et al., 2009). Based on the above mentioned methods it is evident that lots of authors study ionospheric vertical drifts using many different methods. Most of the research and articles are focused on the equatorial or low midlatitudes. A review of the measurements and theoretical models in low latitudes is provided in Fejer (1991). In midlatitudes the ionosphere monitoring is provided by Digisondes and ionosondes. Hence, knowledge of the vertical drifts in the mid-latitudes is still much more limited. 2. Data and methodology In January 2004 a new Digisonde DPS-4 with four cross-loop receiving antennas was installed in the observatory Pruhonice. The Digisonde DPS4 at Pruhonice provides routine vertical ionospheric sounding with a time resolution 15 minutes, except special campaigns of higher sampling rate. The critical frequency (foF2) is automatically detected by an ionogram autoscaling procedure ”Artist” (Galkin et al., 1996, 2008). Data and autoscaled characteristics are distributed to the data centres and presented in real-time on web page http://digisonda.ufa.cas.cz. In addition to the ionogram sounding the drift measurements are performed in both F and E regions. 5
For the F-region drift measurement it is the most convenient to perform an autodrift mode in a short order after ionogram sounding (Reinisch et al., 2005) because the sounding frequencies are automatically selected below the autoscaled foF2 frequency. In our schedule1 the F-region drift measurement follows 5 minutes after the ionogram sounding. The height resolution is set to 5 km and the Doppler shift spectral resolution is 0.195 Hz. Detail description of the measurement is provided in the work Kouba et al. (2008). The output of the measurement in the autoscale mode must be futher manually checked. An employment of the automatic output of the measurement may lead to mixing the reflections from different ionospheric layers during daytime (E and F layers) and during night-time during for instance occurrence of sporadic E layer. Hence, it is important to manually check all the drift data and control them together with the preceding ionogram. The reflection points from raw skymaps were chosen using the selection method described in Kouba et al. (2008). The routine use of the automatically computed velocities may lead to inappropriate smoothing of the effects and/or creation of the artificial effects. In this work we use the reflections from F-region only. The initial drift observations at Pruhonice in 2004 revealed that in the automatic mode measurement it is necessary to control data manually. The DDA method encounters problems especially when range of the measured Doppler frequency shifts is too wide, multiple-hop Es or F reflections are present on ionogram, and velocity distribution across the sounding range is non-uniform Kouba et al. (2008). The automatic drift data processing is significantly improved in the new version of Drift Explorer software (Kozlov and Paznukhov, 2008). However for detailed analysis and research purposes the manual data processing is still inevitable. In this study we analyse data measured in 2006 at Pruhonice station (50 N, 14.6 E). Data set (1 January - 31 December, 2006) represents measurements performed during a period of low solar activity. Maximum value of the solar flux f10.7 cm reached its maximum of 89 sfu (observed) in April. There occurred only 4 short events of high geomagnetic activity classified as a storm events with kP > 5. Such an exceptionally long period of low activity allows us to study the behaviour of the drift velocity during quiet conditions in mid-latitude sector. The number and spatial distribution of detected reflection points relates to 1
described schedule for 2006
6
the current state of the ionosphere. Typically, during the spread F situation the good quality drift data are obtained (Kouba and Kouck´a Kn´ıˇzov´a, 2012). Due to departures from plane stratification the conditions are favourable for registration of large amount of reflected signals. Hence, it allows the accurate velocity estimation. If the ionosphere is stratified into horizontal layers without any spatial disturbances (according to quality of ionogram), we do not detect high number of reflection points necessary for estimating the drift velocity vector. The quantity and the spatial distribution of the detected reflection points mainly affect the accuracy of the horizontal drift components determination. The detected reflection points tend to cumulate near the vertical direction when the ionosphere is horizontally stratified. Such a situation does not allow the estimation of horizontal components. However, when we focus on the vertical velocity component only there is no problem in precise determination. Therefore in this paper we focus on the vertical component of drift velocity vector only. Our database consist of 32 130 (out of 35 040 measurements) manually checked F-region drift velocity vectors. The rest of the measurements were excluded due to the presence of Es layer or ionogram autoscaling failures leading to the incorrect drift measurement. Such a large data set allows us to study the statistical behaviour of velocity vector in quiet conditions. In order to obtain characteristic behaviour of the drift data it is necessary to eliminate influence of gaps with longer duration and/or sudden irregular wave-like events. Except drift velocities directly computed from measurement we use the mean velocities computed with 9 days long window. With this approach we get rid of breaks in the diurnal course of the drift. 3. Results Basic characteristics of the measured vertical velocities are presented in a histogram of these data values (Fig.1). Positive vertical drifts refer to upward plasma motion, while negative vertical drifts refer to downward plasma motion.The preliminary results for the first half of 2006 were published and the quality of the measurements was discussed (Kouba and Kouck´a Kn´ıˇzov´a, 2012). The histogram is well represented by a normal probability density function characterized by the following parameters: median v˜z = −0.35 m s−1 , mean value µ = hvz i = −0.301 m s−1 , and standard deviation σvz = 9.44 m s−1 . 7
Figure 1: Vertical velocity component histogram for F region measurements during 2006.
Someone might infer from the histogram that the vertical drift velocity in mid-latitudinal ionosphere is zero during quiet conditions. Any fluctuation around zero velocity would be caused by irregularities - TID, acoustic-gravity waves, etc. However, the detailed investigation based on the average daily trends does not confirm this expectation.
Figure 2: One-day course of the vertical drift for 2006, Feb 24: a) measured velocities, b) measured velocities smoothed using 30 minutes averaging window, c) averaged velocity measured over 9 days
On the upper panel of Fig.2 there is an example of a one-day-course of the vertical drift component for 2006, Feb 24. Displayed values of the verti8
cal velocity component were calculated for each successfully performed drift measurement without any smoothing of the data. The data show high dispersion. Hence, there is no evident diurnal trend and further process is necessary to derive diurnal course. In general, velocity reaches lower values during the daytime in comparison with the nigh-time values. Tendency of the drift to grow from negative values around sunrise up to positive values close to sunset could be found in the plot. On the middle panel there are the measured velocities displayed as it is typical on the Digisonde web pages. The moving average filter with 30 minutes window is applied on the time series of drift values. The resulting time series is not a smooth line. There are significant breaks and points departuring from the general course. These irregularities occur when gaps, caused by drift measurement failure, are longer than the filter window duration. However, this figure shows a smaller velocities during daytime hours, a higher probability of small negative values in the morning and more positive values in the afternoon. It shows that higher velocity values and stronger wave activity are observed during the night time. Bottom panel of Fig.2 shows the diurnal course of the mean vertical velocity measured for 9 consecutive days. As a representative values for Feb 24 we consider mean values computed for four preceding and four subsequent days (Feb 20 - Feb 28). Comparing the middle and bottom panel we see that both plots are in qualitative agreement. On the bottom plot, there are no discontinuities. Hence the diurnal course is well seen from the plot. The most important feature is that the drift velocity significantly fall to negative values near dawn. The drift velocity reaches its minimum close to the local sunrise. Around the dusk time a similar decrease is observed forming secondary minima of the diurnal course shortly close to sunset. During the daytime, gradual increase of the vertical velocity from negative value after sunrise over the zero around noon to a positive value before sunset is well seen. Averaging groups of nine-days was performed for the data for whole year 2006. On Fig.3 the averaged one-day courses are displayed for the whole year 2006. The following features of vertical velocity diurnal cycles were found: • significant decline (downward peak) around dawn, • less pronounced downward peak around dusk, • increase from small negative values after sunrise over the zero around 9
Figure 3: Average daily trend of the vertical drift for 2006 - seasonal change of the vertical velocity component for 9 days averages
noon to the small positive values before dusk, • no regular course during the night; greater velocity values and higher wave activity. In the diurnal course two regular downward peaks were associated with the dawn and dusk. The existence of the peaks is a general characteristic of all analysed data throughout the year 2006. This fact is evident from Fig.3. Each day the minima of the vertical drift for dawn and dusk were identified on the diurnal course. Seasonal changes of occurrences of the time of these peaks demonstrates Fig.4. It corresponds to the seasonal course of the sunrise and sunset. On the plot, time of the peak minima is emphasised by larger symbols. The beginning of the sudden decrease and the return to normal values before the decrease are marked with smaller symbols. Seasonal course of peaks time occurrences shown on Fig.4 demonstrates linking with dawn and dusk. The peaks are closest to each other in the period around the winter solstice and the farthest apart around the summer solstice. The duration of prevailing negative drift velocities forming these peaks and thus the influence of the dawn/dusk on the drift velocity is mostly 1.5 - 3 hours. It should be noted that around the summer solstice (about May 20 - July 20) the sun does not set for F-region ionospheric heights at the latitude of the Observatory Pruhonice (50 N, 14.9 E). However, the ionization is less effective due to change of the zenith angle. The peaks related to dawn and dusk 10
Figure 4: Seasonal change of the dawn and dusk downward peaks for 9-days averaged daily trend. Large symbols - minima of the peak, small symbols - start of the sudden decrease, return to the values before peak
remain present of the diurnal course. Likely, there is not only a local effect of sunrise and sunset forming the negative velocity structures. The contributing phenomena associated with sunrise and sunset is a passage of the solar terminator. The moving border of the sunlit and and shadow regions forms gravity wave structures important in a global scale (for instance Chimonas and Hines (1970); Somsikov (1991, 1995) ). Gravity wave enhancements associated with motion of the Solar Terminator were reported for instance by ˇ Altadill et al. (2001), Galushko et al. (1998), Cheng et al. (1992), Sauli et al. (2006). Boˇska et al. (2003) found stronger gravity wave activity related to the Solar Terminator in the morning hours. It corresponds to the experimental observation of solar eclipses in the ionosphere. Different atmospheric activity is usually observed during initial and recovery phases of the eclipse ˇ (Jakowski et al., 2008; Sauli et al., 2007) and others) On the plot of mean daily trends Fig.3. velocity enhancements can be seen in the vicinity of the well pronounced minima. However, on the contrary to the well-pronounced negative velocity peaks, the preceding and the following velocity enhancements are difficult to locate. Night-time course of the velocity shows higher day-to-day variability. The post-sunset and presunrise enhancements are not as regular effects as the velocity downward peaks. On the night-time course, the oscillations with periods in the gravity wave domain can be detected/observed. Hence, we focus only on the regular 11
downward velocity peaks.
Figure 5: Seasonal change of magnitudes of decrease for the dawn and dusk peaks; dawn peak - square, dusk peak - triangle
The dawn effect on vertical drift tends to be larger than the effect of the dusk. The observed magnitude of the sunrise and sunset peaks show significant annual course. During first part of the year till solstice, higher variability is evident on both sunrise and sunset peaks. Around solstice, both peaks reaches lowest values. After that, the drift magnitudes of the drift velocity increase smoothly till maxima in summer (August). The vertical drift velocity course is smooth between June solstice and September equinox. The variability increases again after September equinox. The observed wave-like oscillation during winter months within the drift data we attribute to the planetary waves influencing the ionosphere. Most of the observed planetary waves in the ionosphere are propagating from lower atmosphere. Review of the meteorological/lower atmosphere influence on the ionosphere is discussed in details in Laˇstoviˇcka et al. (2006). Much stronger filtering by stratospheric winds in summer is responsible for much smaller penetration of the tropospheric planetary waves to the lower ionosphere heights in summer, which reflects in the observed much smaller summer time planetary wave activity in the ionosphere. This is demonstrated in Fig.5. The magnitudes of decrease for the dawn peaks are marked with squares and for the dusk peak with triangles. The magnitude of the velocity decrease was obtained as a difference of vertical 12
velocity value before the sudden drop and the value in the minimum of the downward peak. Another characteristic feature of all average daily trend of the vertical drift are small velocity values detected during the daytime. Between the dawn and dusk peaks are detected vertical velocities typically within the interval ±10m/s. After the dawn peak recovery there are detected small negative velocity values between −6 to −9 m/s in the most cases. Then the value gradually increases and reaches zero around the local noon. In the afternoon the increase continues till typical values about 6 − 10 m/s before the dusk drop. There were not found any seasonal changes of parameters characterizing this linear increase of vertical velocity during the daytime. We suppose that during day-time during quiet solar and geomagnetic conditions we observe dominant effect of the decrease/increase of the reflection height due to increasing/decreasing zenith angle. Vertical velocities detected during the night hours (between the dusk and the dawn peak) show higher variability compare to day time values. Velocities are generally positive and larger than during the daytime. During night-time the occurrence of wave-like oscillations is a regular characteristics. Typically, the observed waves belong to gravity wave domain as it is seen on the Fig.2. Wave activity is also larger but without any registered regularity. 4. Discussion and Conclusion The drift data of one whole geomagnetically quiet year were processed. This allows us to study the regular behaviour of the drifts in quiet conditions and provides us a base for subsequent studies of the irregular activities during disturbed conditions such as geomagnetic storms, solar eclipses etc. Seasonal course of both sunrise and sunset downward peaks confirm the dominant role of planetary waves in the ionosphere during winter season and around equinoxes. During summer months the course of the sunrise and sunset velocity peaks are smooth without apparent wave-like structures. Apparently, we observe competing mechanisms driving the plasma. We observe plasma motion forced by E×B. The configuration of the geomagnetic field plays important role. The geomagnetic field dip angle is 65.8◦ . The atmospheric waves are another mechanism that significantly influence the plasma in midlatitudes. Within the drift data wave-like structures are well seen.
13
Comparing results from low-latitude ionosphere, the vertical drift in midlatitude differs significantly. In general, the detected values of the observed vertical drift are of lower magnitudes compare to low latitudes. Within data we detect stronger influence of the atmospheric waves than in lower latitudes. On the diurnal course, there is no structure that corresponds to the pre-reversal enhancement as it is observed in low latitudes. The most clearly developed structure is the morning downward velocity peak close to sunrise. This structure occurs regularly and shows clear seasonal dependence. Similar but less pronounced structure, also downward, could be found close to sunset. Smooth increase from sunrise peak values through zero up to positive values preceding the sunset downward peak could be considered as a regular diurnal course. As the measured value of the plasma drift represent the resulting forcing it is difficult to answer what mechanisms is dominant for each moment. We suppose, that during winter season, the plasma motion is forced by the planetary waves. During night-time, the influence of gravity waves is much stronger than during day-time. The dynamics play a more important role in the ionospheric F-region during night-time than during daytime due to the lack of photochemical control at night. During sunrise and sunset downwardpeaks events, the effect of increasing/decreasing ionization contribute to the processes. It coincides with the passage of the Solar Terminator. Hence, the vertical component of the drift reflects the change of the F-layer that could be attributed at least two mechanism. The exact clarification of each contributing mechanism calls for further theoretical study and go beyond the focus of this observational study. Histogram on fig.1 shows that the measured vertical velocity is most frequently in the range of ±20 m/s. Values between 20 and 40 m/s are reported rarely and speeds above 40 m/s can be considered as extreme. It is consistent with the result obtained by Belehaki et al. (2006) where detected amplitude of vertical velocity component seldom exceeds 20 m/s. Similar analysis has been performed by Altadill et al. (2007) on the data measured in 2004 and 2005 in the observatory Ebro (40.8◦ N , 0.5◦ E) by mean of Digisonde 256. The Ebro and Pruhonice stations have one hour difference in local time. Both Ebro and Pruhonice stations belong to mid-latitudes. On the vertical component of the drift velocity gradual increase from negative values close to sunrise to positive values close to sunset has been found in Ebro data as well. This structure is, however, better developed on the diurnal course of the vertical component than in the monthly median courses (Figures 3. and 14
4. in Altadill et al. (2007)). The drop connected with the dawn as a main characteristic of the average daily trend of the vertical drift is well seen in Ebro data. The diurnal course of the vertical component and amplitudes of the detected velocities are in agreement in both stations. Our results are in agreement with published measurements in midlatitudes. We suppose that qualitatively same diurnal course of the drift would be found in all midlatitude stations. In the studied year 2006 there were no significant abnormal ionospheric event/activity ( http://wdc.kugi.kyoto-u.ac.jp/wdc/Sec3.html geomagnetic data, ftp://ftp.swpc.noaa.gov/pub/weekly/RecentIndices.txt solar data) Studied year 2006 was an exceptional year with respect to the solar, geomagnetic and ionospheric activity. There occurred only few isolated events of higher activity. Year 2006 represents last part of the 23rd Solar cycle that ended in 2008. Hence the results of this work may be further used as a reference for drift analyses during special events. This study may be the basis for the evaluation of various phenomena on the drift motion in mid-latitudinal ionosphere. 5. Acknowledgement This work has been supported by Grant Agency of the Czech Republic No. 15-24688S. 6. References References Adebesin, B. O., Adeniyi, J. O., Adimula, I. A., Oladipo, O. A., Olawepo, A. O., Reinisch, B. W., Comparative analysis of nocturnal vertical plasma drift velocities inferred from ground-based ionosonde measurement of hmF2 and h’F. J. of Atmosph. and Sol.-Ter. Phys. 122 (2015), pp. 97107. Adebesin, B.O., Adeniyi, J.O., Adimula, I.A., Reinisch, B.W., Low latitude nighttime ionospheric vertical E B drifts at African region. Advances in Space Research, 52 (2013), pp. 2226-2237 Adeniyi, J. O., Adebesin, B. O., Adimula, I. A., Oladipo, O. A., Olawepo, A. O., Ikubanni, S. O., Reinisch, B. W., Comparison between African 15
equatorial station ground-based inferred vertical E x B drift, Jicamarca direct measured drift, and IRI model. Advances in Space Research, 54(7) (2014), 1629-1641. doi:10.1016/j.asr.2014.06.0 Altadill, D., Sole, J.G., Apostolov, E.M., Vertical structure of a gravity wave like oscillation in the ionosphere generated by the solar eclipse of August 11, 1999. Journal of Geophysical Research 106 (2001), pp. 21419-21428 Altadill, D., Arrazola, D., Blanch, E. F-region vertical drift measurements at Ebro, Spain. Advances in Space Research 39 (2007), pp. 691-698. Batista, I.S., Abdu, M.A., Bittencourt, J.A., Equatorial F region vertical plasma drifts: seasonal and longitudinal asymmetries in American sector. J. Geophys. Res. 91, 12 (1986), pp. 12055-12064 Batista, I.S., de Medeiros, R.T., Abdu, M.A., de Souza, J.R., Bailey, G.J., de Paula, E.R., Equatorial ionospheric vertical plasma drift model over Brazilian region. J. Geophys. Res. 101, 10 (1996), pp. 10887-10892 Belehaki, A., Tsagouri, I., Michalaeras, G., Herekakis, T, F region drift observations from Athens Digisonde. Radio Sci. 41, RS6S35 (2006), doi:10.1029/2005RS003322 Bertoni, F., Batista, I.S., Abdu, M.A., Reinisch, B.W., Kherani, E.A., A comparison of ionospheric vertical drift velocities measured by Digisonde and Incoherent Scatter Radar at magnetic equator. J. of Atmosph. and Sol.-Ter. Phys. 68 (2006), pp. 669-678, doi:10.1016/j.jastp.2006.01.002 Bittencourt, J.A. and Abdu, M.A., A theoretical comparison between apparent and real vertical ionization drift velocities in the equatorial F region. Journal of Geophysical Research 86 (1981), pp. 2451-2454, doi:10.1029/JA086iA04p02451 ˇ Boˇska, J., Sauli, P., Altadill, D., Sol, J. G. Alberca, L. F., Diurnal variation of gravity wave type oscillations at midlatitudes in the ionospheric F region. Studia geophysica et geodaetica 47 (2003), pp. 579-586 Cheng, K., Huang, Y.N., Chen, S.W., Ionospheric effects of the solar eclipse of September 23, 1987, around the equatorial anomaly crest region. Journal of Geophysical Research 97 (1992), pp. 103-111
16
Chimonas G. and Hines C.O., Atmospheric gravity waves induced by a Solar eclipse. J. Geophys. Res., 75 (1970), pp. 875-875 Fejer, B. G., de Paula, E. R., Gonzales, S. A., Woodman, F. R., Average vertical and zonal F-region plasma drifts over Jicamarca, J. Geophys. Res. 96 (1991), pp. 901-906 Fejer., B.G., Low latitude electrodynamic plasma drifts: a review. J. of Atmosph. and Sol.-Ter. Phys. 53, 8 (1991), pp. 677-693 Fejer, B. G., Jensen, J. W., Su, S., Quiet time equatorial F-region vertical plasma drift model derived from ROCSAT-1 observation. J. Geophys. Res. 113 (2008), A05304 Galkin, I. A., Reinisch, B. W., Ososkov, G. A., Zaznobina, E. G., Neshyba, S. P., Feedback neural networks for ARTIST ionogram processing, Radio Sci., 31 (1996), pp. 1119-1129 Galkin, I. A., Khmyrov, G. M., Kozlov, A. V., Reinisch,B. W., Huang, X., Paznukhov,V. V., The ARTIST 5, Radio Sounding and Plasma Physics, AIP Proceedings (2008) Galushko V. G., Paznukhov, V. V., Yampolski, Y. M., Foster, J. C., Incoherent scatter radar observations of AGW/TID events generated by the moving solar terminator. Ann. Geophys. 16 (1998), pp. 821-827. Jakowski, N., Stankov, S. M., Wilken, V., Borries, C., Altadill, D., Chum, ˇ J., Bureˇsova, D., Boˇska, J., Sauli, P., Hruˇska, F., Ionospheric behavior over Europe during the solar eclipse of 3 October 2005. Journal of Atmospheric and Solar-Terrestrial Physics, 70, 6 (2008), pp. 836-853, doi10.1016/j.jastp.2007.02.016 ˇ Kouba, D., Boˇska, J., Galkin, I. A., Santolk, O., Sauli, P., Ionospheric drift measurements: Skymap points selection. Radio Science 43 (2008), RS1S90 Kouba, D., Kouck´a Kn´ıˇzov´a, P., Analysis of digisonde drift measurements quality. Journal of Atmospheric and Solar-Terrestrial Physics, 90-91 (2012), pp. 212-221 Kozlov, A. and Paznukhov, V. V., Digisonde drift analysis software, in Radio Sounding and Plasma Physics, AIP Conf. Proc., 974 (2008), pp. 167-174 17
Kudeki, E., Fawcett, C. D., High resolution observations of 150 km echoes at Jicamarca, Geophys. Res. Lett., 20 (1993), pp. 1987-1990 Laˇstoviˇcka, J., Akmaev, R. A., Beig, G., Bremer, J., Emmert, J. T., Global change in the upper atmosphere. Science, Vol.314, No. 5803 (2006), pp. 1253-1254 Mathew, T. J., Prabhakaran Nayar, S. R., Ravindran, S., Pant, T., Characteristics of the equatorial F-region vertical plasma drift observed during post-sunset and pre-sunrise hours. Advances in Space Research, 46 (2010), pp. 618-625 Prabhakaran Nayar, S. R., Mathew, T. J., Sreehari, C. V., Sumod, S. G., Devasia, et al., Electrodynamics of the equatorial F-region ionosphere during pre-sunrise period. Ann. Geophys., 27 (2009), pp. 107-111 Reinisch, B. W., Scali, J. L., Haines, D. M. Ionospheric drift measurements with ionosondes. Annales Geophysicae 41(5-6) (1998), pp. 695-702 Reinisch, B. W., Huang, X. Galkin, I. A., Paznukhov, V. V., Kozlov, A. V., Recent advances in real-time analysis of ionograms and ionospheric drift measurements with digisondes. Journal of Atmospheric and SolarTerrestrial Physics 67(12) (2005), pp. 1054-1062 Reinisch, B. W., Galkin, I. A., Global Ionospheric Radio Observatory (GIRO), Earth Planets Space, 63 (2011), 377-381 Saranya, P. L., Prasad, D. S. V. V. D., Rama Rao, P. V. S., Ionospheric vertical drifts over an Indian low latitude station and its comparison with IRI-2007 vertical drift model. Advances in Space Research, 54, 6 (2014), pp. 946-954 ˇ Sauli, P., Abry, P., Altadill, D., Boˇska, J., Detection of the wave-like structures in the F-region electron density: Two station measurements. Studia Geophysica et Geodetica, 50, 1 (2006), pp. 131-146, doi:10.1007/s11200006-0007-y ˇ Sauli, P., Roux, S. G., Abry, P., Boˇska, J., Acoustic-gravity waves during solar eclipses: Detection and characterization using wavelet transforms. Journal of Atmospheric and Solar-Terrestrial Physics, 69, 17-18 (2007), pp. 2465-2484, doi:10.1016/j.jastp.2007.06.012 18
Scherliess, L., Fejer, B. G., Radar and satellite global equatorial F region vertical drift model. J. Geophys. Res., 104(A4) (1999), pp. 6829-6842 Somsikov, V. M., Waves in atmosphere excited by Solar Terminator. Geomagnetizm i Aeronomiya, 31 (1991), pp. 1-12 (in Russian) Somsikov, V.M., On the mechanism for the formation of atmospheric irregularities in the solar terminator region. J. Atmos. Terr. Phys., 57 (1995), pp. 75-83 Subbarao, K. S. V., Krishna Murthy, B. V., Post-sunset F region vertical velocity variation at magnetic equator. J. Atmos. Sol. Terr. Phys. 56 (1994), pp. 59-65 Woodman, R. F., Hagfors, T., Method for the measurement of vertical ionospheric motion near the magnetic equator by incoherent scattering, J. Geophys. Res., 75(5) (1969), pp. 1205-1212 Yeh, H. C., Su, S., Heelis, R. A., Wu, J., The ROCSAT-1 IPEI Preliminary Results: Vertical Ion Drift Statistics. TAO, 10 (4) (1999), pp. 805-820
19