Revised comparison and analysis of severe magnetic storm effects measured over Istanbul and Rome

Geomagnetic Storms (GS) and Ionospheric Storms (IS) for high geomagnetic activity conditions are studied in Istanbul (IST) and in Rome (RO). Further and revised results that led to new interpretations are briefly presented here following a previous paper on the severe storm effects on the ionosphere over Istanbul. The series of storms started on 28th October 1968 and went on to 2nd November 1968 with the geomagnetic index Kp reaching values up to 8. Deviations of the critical frequency of the F2 layer, ∆foF2(%), from monthly medians up to 100% were observed. Power spectral results show here that not only the Diurnal Planetary Waves (PW) of 2.3 day periodicity but also tidal waves with semidiurnal (12h), teridiurnal (8h) and quaterdiurnal (6h) characteristic times are present. Mailing address: Dr. Harutyun Agopyan, Istanbul University, Vocational School of Technical Sciences, Telecommunication Program, 34320 Avcilar, Istanbul, Turkey: e-mail: harutyun@istanbul.edu.tr Vol52,5,2009 1-12-2009 13:29 Pagina 453

A similarity between the two stations situated at the same L shell (1.6) and with only ∆h=1h difference of universal standard time (UTS) as well a slightly different behavior typical of disturbed days, can also be clearly seen in fig. 2.
The plasmapause is typically located at L=4-6 but may be found at L ~ 2 during very active periods; for both observation sites at L=1.6 and with (magnetic planetary index Kp) Kp ≥7 then the saturation of the protonospheric reservoir value should take (see Kamide and Richmond, 1986) ts = 0.17(L) 4 = _ 1.1 days, so that one day becomes insufficient for its replenishment as seen in fig. 3.
Magnetic recordings of GSs and ionospheric soundings for ISs from 28th October 1968 to 2nd November 1968 with geomagnetic 3hourly magnetic activity index Kp reaching up to 7 to 8 values are considered.During the month of October H mean was -16.1 nT but in November the H mean was = -32.5 nT (around twice of the annual mean value = -16.1 nT).
The geomagnetic field disturbances during this time interval are reported in Agopyan, 2002, figure 6.The maximum disturbance in H component is reached on 31st October around 21UT.In the first storm day the ionospheric height changes are well over 250 km.Leitinger et al., (2004) give slab thickness (τ) values at midday as 320 km and at midnight as 340 km over RO using SIRO data during October 1981.Whereas on 29th October IST measured NmF2= 7x10 11 e/m 3 , Zhang et al., (2003) give values for RO as NmF2=1x10 -6 /cm -3 for 15th July 2000.
Using new IRI 2007 model (http://IRI.gsfc.nasa.gov2008) no changes in height occurs during sunlight hours but electron density (Ne) changes are very significant during 18LT hrs for RO and 19LT hrs for IST simultaneously.At h=300 km the maximum electron density (Nm) of RO (Nm=8x10 11 m -3 ) shows approximately ∆Nm=4x10 11 m -3 concentration difference as max According to the reports by Prölss and Werner (2002) the largest storm effects have been observed at RO.By using the IRI-2000 ionospheric model, applied to RO data, a high degree of reliability during quiet conditions and inadequate degree of reliability during storm events has been shown (Cander, 2003;Zolesi and Cander, 2004).

Results of comparisons
The percentage frequency (MHz) deviations ∆foF2 (%) reached up to 70%.The comparison plots of normalized prestorm analysis in fig. 2 showed a deviation of about ± 2%.In both ionospheric stations the maximum electron densities were approximately Nm=1.45x10 6 e/cc, while the heights of the maximum were about h=280 km.
Figure 3 shows the percentage deviation of foF2 (MHz) during the magnetic storm period of 5 days.
Planetary waves (periods of about 2-30 days) were predominantly of tropospheric origin and penetrated directly to heights slightly above 100 km.They have to propagate upwards into the F-region ionosphere via an indirect route.Their effects were observed in the lower ionosphere, in the ionospheric E region in h'E and sporadic-E (Es) layer and in the F2 region (e.g., Altadill and Apostolov, 2001;Lastovicka et al., 2006;Mikhailov et al., 2004).The quasi-2-day wave in the Mesosphere Lower Thermosphere (MLT) was the most studied.Summarizing, it has been reported that the 2-day wave is a manifestation of both a global normal Rossby mode and a local unstable wave.Its activity is maximum during local summer and especially close to equinoxes (e.g., Altadill and Apostolov, 2001).Tides in the atmosphere are predominantly thermal, not gravitational tides.They are generated by the periodic solar heating.Dominant tidal periods in the ionosphere are 24 h and 12 h.The spectral analysis was applied to the Rome and Istanbul data.The power spectra including the Hanning windowing showed the effects of planetary waves (PW) with peaks at 2.3 days and tidal waves with diurnal and semidiurnal peaks, fig. 4.
It is also notable that during the negative phases of the IS, ∆foF2 minimum value mostly restricting telecommunication channels and radio propagations observed over IST (∆foF2=-33%) was 10% greater than that observed over RO (∆foF2=-22%); whereas during the positive phases of the IS ∆foF2=+70% was observed over IST while in RO ∆foF2= was +75%.

Conclusions
Results of the analysis of the studied severe GSs presented here show complex mechanisms for the ionosphere at medium range scale as also partly seen in previous papers on IST using storm sudden commencements (SC) and ionospheric storms (IS) (Agopyan, 1986 and2002).In this paper a comparison between RO and IST for the power spectral analysis on foF2 data showed peaks at 2.3 days and at 24h (diurnal), 12h (semidiurnal), 8h (teridiurnal) and 6h (quaterdiurnal) characteristics in both stations.The peaks magnitude was however weaker in RO.
The morphological analysis of the two stations magnetic and ionospheric effect can be now compared and summarized as follows: -FIRST STORM (1st SC at 09h09 UT on 29th October 1968, Kp=7): After the sudden commencement following weak ring currents, a rapid decrease of ∆H and increase of ∆foF2 on 30th October 1968 at 02h LT in IST, and 00h LT in RO, was observed simultaneously.
-SECOND STORM (2nd SC at 09h00 UT on 31st October 1968, Kp=8): A very strong negative phase for ∆H started immediately after the SC indicating the strong effect of the ring current increase; the foF2 data show small excursions following SCs exhibiting positive phase for RO but negative for IST.The largest positive response came with a delay about 8 hs for IST but 12 hs for RO with respect to SC.
-THIRD STORM (3rd SC at 09h17 UT on 1st November 1968, Kp=7): An apparent ∆H peak, with the rapid increase and decrease in negative phase which proves recovery of the intensive ring currents coinciding with a positive phase in ∆foF2 and an elevation of the F layer for both stations IST and RO were observed consecutively.The ionospheric positive and negative wavelike fluctuations were less significant in RO than in IST (Agopyan, 1994).Negative phases last for a long time, even several hours after a SC and were observed in all three cases.Negative long-duration F2-layer storms are known to result from thermospheric neutral composition variations, namely, O/N2 decrease (e.g., Prölss, 1995).Vol52 ,5,2009,5, 1-12-2009 13:29 Pagina 457 13:29 Pagina 457

Fig. 1 .
Fig. 1.Comparisons of the ionospheric heights (km) versus electron density (Ne/cc) generated noon profiles of IRI for Istanbul solid-line (stars) and Rome broken-line (circles) plotted simultaneously.

Fig. 3 .
Fig. 3. Comparisons of percentage deviations of foF2 (MHz) from monthly medians versus time of 5 days after 29 October to 2 November 1968 for Istanbul solid-line (stars) and Rome broken-line (circles) plotted simultaneously.

Fig. 4 .
Fig. 4. Comparison results of power spectra of electron density values of the ionospheric severe storm including Hanning windowing for Istanbul (black-line) and Rome (dashed black-line) plotted simultaneously.