Vol50,3,2007 483 ANNALS OF GEOPHYSICS, VOL. 50, N. 3, June 2007 Key words ionosonde – ionospheric data – auto- scaling – equatorial anomaly 1. Introduction Since the early years of the last century, ionospheric measurements have played a signifi- cant role in studies concerning ionospheric physics and related phenomena. By analyzing an ionogram important ionospheric characteristics can be extracted, that can strongly contribute to the knowledge of physical phenomena such as ra- dio propagation in ionized media, physical and chemical processes in upper atmosphere, iono- sphere and magnetosphere coupling, and solar- terrestrial relations. Furthermore, observation of ionospheric plasma and forecasting of physical phenomena connected to the Sun-Magnetos- phere-Ionosphere-Thermosphere system holds a remarkable scientific interest with respect to Space Weather, because of the influence of these phenomena on satellite and terrestrial communi- cations. For such a purpose ionospheric sounders should have some distinctive features, especial- ly oriented towards routine service, like Internet connection and the possibility to perform an au- tomatic scaling of the recorded ionograms in re- al time. The Advanced Ionospheric Sounder (Aroki- asamy et al., 2002; Zuccheretti et al., 2003), built at the Istituto Nazionale di Geofisica e Vulcanologia, Rome, Italy (here abbreviated AIS-INGV) satisfies these requirements. Like other recent sounders, AIS-INGV iono- sonde is practically built around a PC and has been designed to fulfill certain physical charac- teristics such as the power reduction (around 200 W against several kilowatts of traditional sys- The new ionospheric station of Tucumán: first results Michael Pezzopane (1), Enrico Zuccheretti (1), Cesidio Bianchi (1), Carlo Scotto (1), Bruno Zolesi (1), Miguel A. Cabrera (2)(3) and Rodolfo G. Ezquer (2)(3)(4) (1) Istituto Nazionale di Geofisica e Vulcanologia, Roma, Italy (2) CIASUR, Facultad Regional Tucumán, Universidad Tecnológica Nacional, Tucumán, Argentina (3) Laboratorio de Ionósfera, Instituto de Física, Universidad Nacional de Tucumán, Argentina (4) Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET), Buenos Aires, Argentina Abstract An Advanced Ionospheric Sounder, built at the Istituto Nazionale di Geofisica e Vulcanologia, Rome, Italy, was installed at Tucumán, Argentina, particularly interesting for its location, near the southern peak of the ionospher- ic equatorial anomaly. The aim of this installation is to collect a large number of continuous data useful both to study the dynamics of the equatorial ionospheric plasma and to develop reliable regional ionospheric prediction models. Moreover this ionosonde will contribute to the ionospheric database and real time knowledge of South- ern Hemisphere ionospheric conditions for space weather applications. The ionosonde is completely program- mable and two PCs support the data acquisition, control, storage and on-line processing. In this work the first results, in terms of ionograms and autoscaled characteristics, are presented and briefly discussed. Mailing address: Dr. Michael Pezzopane, Istituto Na- zionale di Geofisica e Vulcanologia, Via di Vigna Murata 605, 00143 Roma, Italy; e-mail: pezzopane@ingv.it 484 Michael Pezzopane, Enrico Zuccheretti, Cesidio Bianchi, Carlo Scotto, Bruno Zolesi, Miguel A. Cabrera and Rodolfo G. Ezquer tems) and consequently weight, size, power con- sumption and hardware complexity. It exploits the computer resources to manage the sounding, the real time signal processing, data storing and sharing; it has also the capability to be remotely programmable. The basic work of this ionosonde is to generate an ionogram from which virtual heights and critical frequencies can be scaled. But this ionosonde is also equipped with a soft- ware, called Autoscala (Scotto and Pezzopane, 2002), able to automatically scale in real time from an ionogram the following characteristics: foF2, MUF(3000)F2, M(3000)F2, fxI, foF1, ftEs, and h’Es. Within the Italian-Argentine collaboration supported by the Istituto Italo Latino Americano (IILA), an AIS-INGV ionosonde was installed at Tucumán (geographical coordinates: 26.9S, 294.6E; magnetic coordinates: 15.5S, 3.8E), Ar- gentina (fig. 1), near the ionospheric southern Equatorial Anomaly (EA) crest, at the end of Au- gust 2007. From 1957 to 1987 a first generation analog ionosonde operated at the same location contributing to the knowledge of the South Amer- ican ionosphere. However the characteristics of the new ionospheric station are such that the data obtained can contribute significantly both to the study of the dynamics of the low latitudes iono- sphere (for instance in terms of the occurrence of the equatorial spread-F, whose onset is strictly re- lated to the EA development) and to built up reli- able regional ionospheric prediction models for radio communications. This paper briefly describes the main char- acteristics of the AIS-INGV/Autoscala system installed at Tucumán, and then it illustrates the first results of the operating ionospheric station, in terms of recorded ionograms and autoscaled characteristics. 2. The AIS-INGV/Autoscala system: a brief description The aim of this ionosonde is to fulfil the de- mand to have a simple low-cost system to sound the ionosphere. To do this a system has been designed in which the most advanced HF radar techniques have been employed (Skolnik, 1981, 1990). In order to compensate the power reduction, a 16 bit complementary phase code is employed (Golay, 1961) together with pulse-compression and a phase coherent integration, giving the possibility to investigate the ionosphere with a power of 200 W only. The ionosonde is com- Fig. 1. Geographical location of the Tucumán ionospheric station. 485 The new ionospheric station of Tucumán: first results pletely programmable and the data acquisition, control, storage and on-line processing are sup- ported by a PC (Arokiasamy et al., 2002; Bianchi et al., 2003; Zuccheretti et al., 2003). The antenna system is a typical crossed delta configuration in which a single backstayed mast sustains the radiative elements. They are 90° ori- ented to limit the energy cross talk between transmitting and receiving antennas. The trans- mitting and receiving antennas have been built by engineers from the Facultad Regional Tu- cumán (FRT) of the Universidad Tecnológica Nacional (UTN); they are triangle shaped, 21 m high and 40 m long, fed by baluns and loaded by proper resistors. The loads are made up with spe- cial non inductive resistors to widen the band and to better direct the lobe upwards; the baluns create the proper match with 50 Ω lines and power amplifier. The product of the sounding process is a raw ionogram that is made available to be archived as well as to be processed real time by Autoscala operating in another PC (fig. 2). The characteris- tics given automatically as output by Autoscala are foF2, MUF(3000)F2, M(3000)F2, fxI, foF1, ftEs, and h’Es. Autoscala is based on an image recognition technique, it does not use information on polar- ization, and it can then be applied to any kind of antenna system. Another important characteris- tic of Autoscala is its behaviour for ionograms characterized by a truncated trace; in these cas- es, if the digital information of the ionogram is considered sufficient, the software can recon- struct the absent trace, giving more reliable au- tomatically scaled values. Since the first phase of Autoscala develop- ment attention has been paid to the quantitative evaluation of the performance of the algorithms by comparing the output from Autoscala with the corresponding result obtained by a well ex- perienced operator. These tests have demon- strated the reliability of the software both for quiet and for disturbed conditions (Pezzopane and Scotto, 2004, 2005). Firstly Autoscala tries to identify on the ionogram the F2 layer. In order to do this, the ionogram is initially memorized as a matrix. Then two empirical curves, that are able to fit the typical shape of the F2 trace, and character- ized by several parameters, are defined (Pez- zopane and Scotto, 2007a). For each set of curves the local correlation C with the recorded ionogram is calculated making allowance for both the number of matched points and their amplitude. The set of curves having the maxi- Fig. 2. The AIS-INGV/Autoscala system data flow. 486 Michael Pezzopane, Enrico Zuccheretti, Cesidio Bianchi, Carlo Scotto, Bruno Zolesi, Miguel A. Cabrera and Rodolfo G. Ezquer mum value of C is then selected. If this value of C is greater than a fixed threshold Ct the select- ed curves are considered as representative of the F2 trace. foF2 and MUF(3000)F2 are then ob- tained from the selected curve representing the ordinary ray, foF2 as the frequency of the verti- cal asymptote, while MUF(3000)F2 is numeri- cally calculated by finding the corresponding tangent transmission curve. fxI is obtained as the frequency of the vertical asymptote of the se- lected curve representing the F2 extraordinary ray. M(3000)F2 is simply the ratio between the calculated MUF(3000)F2 and foF2. On the con- trary if C does not exceed Ct the F2 layer au- toscaling routine considers the digital information of the ionogram not sufficient to establish whether the F2 trace is visible on the ionogram, and the output for foF2, MUF(3000)F2, M(3000)F2, and fxI will be N/A (Not Available). The algorithm developed to identify the F1 layer automatically (Pezzopane and Scotto, 2007b) is very similar to the one used for scaling the F2 layer. The main difference is that unlike the F2 layer autoscaling procedure, which is based on the identification of both ordinary and extraordi- nary rays, this procedure tries to fit only the F1 or- dinary ray by using a set of parabolas. For each parabola the local correlation C with the recorded ionogram is calculated making allowance for both the number of matched points and their ampli- tude. The parabola having the maximum value of C is then selected. If this value of C is greater than a fixed threshold Ct, the selected parabola is con- sidered representative of the F1 ordinary ray. Once the parabola considered as representative of the F1 ordinary trace is selected, foF1 corre- sponds with the frequency of the point of this parabola which has the highest virtual height. On the contrary, if C does not exceed Ct two alterna- tive outputs for foF1 are possible: NO if the F1 cusp is not observed, or N/A if the digital infor- mation of the ionogram is considered not suffi- cient to establish whether the F1 cusp is present or not. It is important to point out that the F1 layer autoscaling routine is not able to function if the F2 layer has not been identified. This because the al- gorithm needs as input parameter the minimum virtual height of the F2 ordinary trace. In addition, the routine to automatically scale the E sporadic (Es) layer (Scotto and Pez- zopane, 2007) has been developed along similar lines to the F2 and F1 layer routines. If only the ordinary component is recorded the output cor- responds to foEs, while if only the extraordi- nary component is recorded the output corre- sponds to fxEs. The ionograms recorded by AIS-INGV show both components and as a consequence the Es routine limits itself to give as output ftEs (defined as the top frequency of the Es layer according to Wakai et al., 1987) be- ing unable to specify this value as foEs or fxEs. The technique relies on a set of curves having the typical shape of the Es layer. Appropriate bounds for the height and the frequency are set. In particular curves having maximum frequen- cy lower than the modelled critical frequency foE of the normal E region are not considered. For each curve the local correlation C with the recorded ionogram is calculated with allowance made for both the number of matched points and their amplitude. The curve having the max- imum value of C is then selected. If this value of C is greater than a fixed threshold Ct the se- lected curve is considered representative of the Es trace. The value ftEs is thus obtained as the maximum frequency of the curve together with the associated height h’Es. On the contrary if Ct is not exceeded then the routine assumes the Es trace is not present on the ionogram and the output for ftEs and h’Es will be NO. The processing time of Autoscala is approxi- mately 50 s on a computer with 1.60 GHz proces- sor and 512 MB of RAM. 3. The ionospheric station of Tucumán: its importance and first results The EA is characterized by two enhanced plasma crests on both sides of the magnetic equa- tor at about ±20° magnetic latitude. The electro- dynamic drift theory has successfully explained the main features of the EA. According to this theory, north-south geomagnetic field combined with the day-time east-west ionospheric electric field (both fields parallel to the surface of the Earth at the equator) creates a plasma fountain rising up to several hundred kilometers. This up- ward drifting plasma, when losing its momen- tum, it moves under gravity along the geomagnet- 487 The new ionospheric station of Tucumán: first results ic lines to higher latitudes creating the crests. This fountain effect and the consequent anomaly may encompass more than 30° latitude on either side of the magnetic equator. The plasma fountain and the anomaly exhibit north-south asymmetries with respect to the geomagnetic equator mainly due to field-aligned plasma flows caused by neu- tral winds (Titheridge, 1995). Balan and Bailey (1995) have studied the equatorial anomaly over Jicamarca (geographical coordinates: 12S, 76.8W; magnetic coordinates: 2.4S, 4.9W) using the Sheffield University plasmasphere-ionosphere model and they have shown the possible exis- tence of an additional layer, first called the G lay- Fig. 3. An example of a nighttime ionogram recorded on 28 September 2007 at 00:20 UT by the AIS-INGV ionosonde installed at Tucumán, and autoscaled by Autoscala. Fig. 4. An example of a daytime ionogram recorded on 27 September 2007 at 12:55 UT by the AIS-INGV ionosonde installed at Tucumán, and autoscaled by Autoscala. 488 Michael Pezzopane, Enrico Zuccheretti, Cesidio Bianchi, Carlo Scotto, Bruno Zolesi, Miguel A. Cabrera and Rodolfo G. Ezquer F ig . 5. Io no gr am s re co rd ed o n 23 S ep te m be r 20 07 f ro m 1 4: 05 t o 14 :4 5 U T b y th e A IS -I N G V i on os on de i ns ta ll ed a t T uc um án , an d au to sc al ed b y A ut os ca la . In a ll b ut t he f ir st a nd l as t, de ve lo pm en t an d de ca y of a F 1. 5 ad di ti on al s tr at if ic at io n ar e hi gh li gh te d us in g op en c ir cl es . 489 The new ionospheric station of Tucumán: first results er, and now renamed the F3 layer. Features and occurrence of this additional stratification have been observed and studied by Rama Rao et al. (2005), and Batista et al. (2002). Thampi et al. (2007) have also observed signatures of this ad- ditional layer in the latitudinal profiles of Total Electron Content (TEC). Additional stratifica- tions occur also below the foF2 at low latitudes as reported by Lynn et al. (2000). These stratifi- cations that do not alter the F2 maximum conti- nuity are called F1.5. In the light of all these re- cent studies, the location of the ionospheric sta- tion of Tucumán assumes a great importance be- cause of the very limited number of operating ionospheric stations in South America, and also because this new ionospheric station is potential- ly able to generate and store a large number of continuous data. Figures 3, 4, and 5 show examples of iono- grams respectively nighttime, daytime, and with Fig. 6. Hourly foF2 plots from 07 to 12 September 2007 as obtained by the ionograms recorded by the AIS- INGV ionosonde installed at Tucumán. Values manually scaled, and values obtained automatically by Autoscala, are indicated by open circles, and solid squares, respectively. 490 Michael Pezzopane, Enrico Zuccheretti, Cesidio Bianchi, Carlo Scotto, Bruno Zolesi, Miguel A. Cabrera and Rodolfo G. Ezquer possible additional stratifications, recorded by the AIS-INGV/Autoscala system installed at the new ionospheric station of Tucumán. Although the ordinary and the extraordinary rays are very close, Autoscala succeeds in giv- ing as output reliable automatically scaled val- ues. As evidence of the reliability of the au- toscaled values figs. 6 and 7 illustrate two se- quences of six days, from 07 to 12 September 2007 (Kp = 4+), and from 29 September to 04 October 2007 (Kp = 5), for which the foF2 val- ues obtained manually are compared with the corresponding ones scaled by Autoscala. Even though both sequences of days are geomagnet- ically quiet, unlike the second one the first se- quence presents a high day-to-day variability, typical at low latitudes, that is well matched by Autoscala. This points out that a long series of autoscaled data obtained at Tucumán can pro- vide a valuable database for studying the day- Fig. 7. Hourly foF2 plots from 29 September 2007 to 04 October 2007 as obtained by the ionograms recorded by the AIS-INGV ionosonde installed at Tucumán. Values manually scaled, and values obtained automatically by Autoscala, are indicated by open circles, and solid squares, respectively. 491 The new ionospheric station of Tucumán: first results to-day variability of the equatorial plasma, both for quiet and for disturbed conditions, which is to be still completely understood. The same database may also be used to develop reliable regional ionospheric prediction models. Moreover, analysis of the autoscaled validat- ed data has also led to interesting conclusions concerning the importance that the sounding rep- etition rate of the ionosonde has at these latitudes. In fact during the testing phase it has been seen that an hourly repetition rate could hide interest- ing phenomena well visible on the contrary with a higher repetition rate, as shown in fig. 8a,b. 4. Summary This work describes the installation of a new ionospheric station at Tucumán, Argentina, par- ticularly interesting for its location, exactly in the southern equatorial anomaly. The station was equipped with an AIS-INGV/Autoscala system, able to give as output autoscaled ionos- pheric values. For this reason the station can contribute to the ionospheric database and can be a part of a possible net for space weather pur- poses. On this subject it is worth noting that the real time foF2 values produced by the station are already being used by the Australian IPS Radio and Space Services for mapping purposes (see the site ). To date, the ionograms recorded at the Tu- cumán ionospheric station by the ionosonde AIS- INGV, and autoscaled by Autoscala, are accessi- ble real time through the site . Acknowledgements The authors gratefully acknowledge Prof. Sandro Maria Radicella for his helpful sugges- tions. REFERENCES AROKIASAMY, B J., C. BIANCHI, U. SCIACCA, G. TUTONE and E. ZUCCHERETTI (2002): The new AIS-INGV digital ionosonde design report, INGV Int. Tech. Rep. No. 12. BALAN, N. and G.J. BAILEY (1995): Equatorial plasma foun- tain and its effects: Possibility of an additional layer, J. Geophys. Res., 100 (A11), 21,421-21,432. BATISTA, I.S., M.A. ABDU, J. MACDOUGALL and J.R. SOUZA (2002): Long term trends in the frequency of occur- rence of the F3 layer over Fortaleza, Brazil, J. Atmos. Sol. Terr. Phys., 64 (12-14), 1409-1412. BIANCHI, C., U. SCIACCA, A. ZIRIZZOTTI, E. ZUCCHERETTI and B.J. AROKIASAMY (2003): Signal processing for phase-coded HF-VHF radars, Ann. Geophysics, 46 (4), 697-705. GOLAY, M.J.E. (1961): Complementary series, IRE Trans. Inform. Theory, 7, 82-87. LYNN, K.J.W., T.J. HARRIS and M. SJARIFUDIN (2000): Strat- ification of the F2 layer observed in Southeast Asia, J. Geophys. Res., 105 (A12), 27,147-27,156. PEZZOPANE, M. and C. SCOTTO (2004): Software for the automatic scaling of critical frequency foF2 and Fig. 8a,b. a) Hourly and b) 5 min foF2 plots of 01 October 2007 as obtained by the ionograms recorded by the AIS-INGV/Autoscala system installed at Tucumán. In this case an hourly repetition rate hides an interesting phe- nomenon, from 05:00 to 08:00 UT, well visible on the contrary with a higher repetition rate. a b 492 Michael Pezzopane, Enrico Zuccheretti, Cesidio Bianchi, Carlo Scotto, Bruno Zolesi, Miguel A. Cabrera and Rodolfo G. Ezquer MUF(3000)F2 from ionograms applied at the ionos- pheric observatory of Gibilmanna, Ann. Geophysics, 47 (6), 1783-1790. PEZZOPANE, M. and C. SCOTTO (2005): The INGV software for the automatic scaling of foF2 and MUF(3000)F2 from ionograms: a performance comparison with AR- TIST 4.01 from Rome data, J. Atmos. Sol. Terr. Phys., 67 (12), 1063-1073. PEZZOPANE, M. and C. SCOTTO (2007a): The automatic scal- ing of critical frequency foF2 and MUF(3000)F2: a comparison between Autoscala and ARTIST 4.5 on Rome data, Radio Sci., 42, RS4003, doi: 10.1029/ 2006RS003581. PEZZOPANE, M. and C. SCOTTO (2007b): A method for auto- matic scaling of F1 critical frequencies from iono- grams, Radio Sci. (in press). RAMA RAO, P.V.S., K. NIRANJAN, D.S.V.V.D. PRASAD, P.S. BRAHMANANDAM and S. GOPIKRISHNA (2005): Features of additional stratification in ionospheric F2 layer observed for half a solar cycle over Indian low latitudes, J. Geo- phys. Res., 110 (A04307), doi: 10.1029/2004JA010646. SCOTTO, C. and M. PEZZOPANE (2002): A software for auto- matic scaling of foF2 and MUF(3000)F2 from iono- grams, in Proceedings of URSI 2002, 17-24 August 2002, Maastricht (paper 1018 of the electronic copy). SCOTTO, C. and M. PEZZOPANE (2007): A method for auto- matic scaling of sporadic E layers from ionograms, Ra- dio Sci., 42 (RS2012), doi: 10.129/2006RS003461 SKOLNIK, M.I. (1981): Introduction to Radar Systems (Mc- Graw Hill, New York), pp. 581. SKOLNIK, M.I. (1990): Radar Handbook (McGraw Hill, New York), pp. 1232. THAMPI, S.V., N. BALAN, S. RAVINDRAN, T.K. PANT, C.V. DEVASIA, P. SREELATHA, R. SRIDHARAN and G.J. BAILEY (2007): An additional layer in the low-latitude ionosphere in Indian longitudes: total electron content observations and modelling, J. Geophys. Res., 112 (A06301), doi: 10.1029/2006JA011974. TITHERIDGE, J.E. (1995): Winds in the ionosphere-A review, J. Atmos. Sol. Terr. Phys., 57 (14), 1681-1714. WAKAI, N., H. OHYAMA and T. KOIZUMI (1987): Manual of Ionogram Scaling, 3rd version (Radio Research Labora- tory Ministry of Posts and Telecommunications, Japan), pp. 119. ZUCCHERETTI, E., C. BIANCHI, U. SCIACCA, G. TUTONE and B.J. AROKIASAMY (2003): The new AIS-INGV digital ionosonde, Ann. Geophysics, 46 (4), 647-659. (received September 24, 2007; accepted October 29, 2007)