The International Reference Ionosphere – Status 2013

Available online at www.sciencedirect.com

ScienceDirect Advances in Space Research xxx (2014) xxx–xxx www.elsevier.com/locate/asr

The International Reference Ionosphere – Status 2013 Dieter Bilitza ⇑ George Mason University, School of Physics Astronomy and Computational Science, 4400 University Drive, Fairfax, VA 22030, USA NASA Goddard Space Flight Center, Heliospheric Physics Laboratory, Greenbelt, MD 20771, USA Received 26 March 2014; received in revised form 24 July 2014; accepted 25 July 2014

Abstract This paper describes the latest version of the International Reference Ionosphere (IRI) model. IRI-2012 includes new models for the electron density and ion densities in the region below the F-peak, a storm-time model for the auroral E-region, an improved electron temperature model that includes variations with solar activity, and for the first time a description of auroral boundaries. In addition, the thermosphere model required for baseline neutral densities and temperatures was upgraded from MSIS-86 to the newer NRLMSIS-00 model and Corrected Geomagnetic coordinates (CGM) were included in IRI as an additional coordinate system for a better representation of auroral and polar latitudes. Ongoing IRI activities towards the inclusion of an improved model for the F2 peak height hmF2 are discussed as are efforts to develop a “Real-Time IRI”. The paper is based on an IRI status report presented at the 2013 IRI Workshop in Olsztyn, Poland. The IRI homepage is at IRImodel.org. Ó 2014 COSPAR. Published by Elsevier Ltd. All rights reserved.

Keywords: Ionosphere; IRI; Forecast; Space Weather; Real-Time

1. Introduction The International Reference Ionosphere (IRI) is an international project that was initiated by the Committee on Space Research (COSPAR) and the International Union of Radio Science (URSI) with the goal of establishing a standard representation of the plasma parameters in Earth’s ionosphere. Such a model of the ionosphere is important for the many applications that rely on electromagnetic waves traveling through the ionosphere including telecommunication, GPS, earth observation from space (e.g., satellite altimetry), radio astronomy and many more, because all of these applications need to correct for the retarding and refractive effect of the ionosphere on the probing signal. At the request of COSPAR and URSI, ⇑ Address: George Mason University, School of Physics Astronomy and Computational Science, 4400 University Drive, Fairfax, VA 22030, USA. Tel.: +1 301 922 4769; fax: +1 301 286 0190. E-mail addresses: [email protected], [email protected].

IRI is an empirical model being based on the majority of the available and reliable ground and space observations and avoiding as much as possible dependence on the still evolving theoretical understanding. But theoretical considerations can be helpful in bridging data gaps and for internal consistency checks. COSPAR’s prime interest in IRI is as a general description of the ionosphere as part of the terrestrial environment for the evaluation of environmental effects on spacecraft and experiments in space. URSI’s prime interest is in the electron density part of IRI for defining the background ionosphere for radiowave propagation studies and applications. With COSPAR and URSI the IRI project has the backing from the major international unions representing space-based ionospheric observations (COSPAR) and ground-based ionospheric observations (URSI). IRI development has benefitted greatly from the synergism between these two communities that are represented about evenly in the IRI Working Group and during bi-annual IRI Workshops. The current membership roster of the IRI Working Group is listed in

http://dx.doi.org/10.1016/j.asr.2014.07.032 0273-1177/Ó 2014 COSPAR. Published by Elsevier Ltd. All rights reserved.

Please cite this article in press as: Bilitza, D. The International Reference Ionosphere – Status 2013. J. Adv. Space Res. (2014), http://dx.doi.org/ 10.1016/j.asr.2014.07.032

2

D. Bilitza / Advances in Space Research xxx (2014) xxx–xxx

Table 1. Progress of the IRI project is documented in several dedicated issues of the journal Advances in Space Research with selected papers from these workshops (see http://irimodel.org/docs/asr_list.html for references). The model has undergone a continuous improvement process that resulted in the release of major new editions of the model about every five years (Rawer et al., 1975; Rawer et al., 1978; Rawer et al., 1978; Rawer et al., 1981; Bilitza, 1986; Bilitza, 1990; Bilitza, 1997; Bilitza, 2001; Bilitza and Reinisch, 2008). An important goal of geospace science is to develop predictions and forecast capabilities in support of human presence and technical endeavors in geospace. Data-based models like IRI are an important element of this undertaking because they compress the large volume of observational evidence in mathematical form combining different data sources and different techniques. One particular advantage over theoretical models is the very fact that they do not depend on the evolving theoretical understanding of the heliospheric environment. There are several examples where effects were already included in IRI before they were fully understood and included in theoretical models. One example is the distinct longitudinal variation with 4 maxima (wave number 4 structure) of the F-peak electron density that was first reported by Benkova et al. (1990) based on Interkosmos 19 topside sounder data and later with IMAGE/EUV observations (Immel et al., 2006), and then

confirmed with data from CHAMP (Lu¨hr et al., 2007) and TOPEX (Scherliess et al., 2008). While theoretical models still struggle to include this phenomenon in their modeling framework, IRI already includes a smoothed version of this phenomenon (McNamara et al., 2010); a smoothing effect is to be expected since IRI is based on monthly averages. Other examples are the midlatitude evening anomaly (Weddell Sea anomaly) observed by the Super Dual Auroral Radar Network (SuperDARN) radars that is well captured by IRI (de Larquier et al., 2011) and the occurrence of ionospheric plasma caves under the Equatorial Ionization Anomaly (EIA) (Lee et al., 2012). A disadvantage of empirical models is the strong dependence on the underlying data base. Regions and time periods not well covered by the data base will result in diminished reliability of the model in these areas. So, for example, conditions during the most recent solar minimum in 2008/2009 were very different from earlier minima. The minimum was lower and more extended than earlier minima and as a result IRI being built with the data from earlier minima overestimated the plasma densities during the minimum period (Lu¨hr and Xiong, 2010; Bilitza et al., 2012). In this article we will first present the changes that led to IRI-2012 version of the model and then will discuss two important activities towards future improvements: a better representation of the F peak height hmF2 and the development of a Real-Time IRI.

Table 1 IRI Working Group members and steering committee.

Steering Committee: L.-A. McKinnell (South Africa) Chair V. Truhlik (Czech Rep.) URSI Vice-Chair, S. Watanabe (Japan) COSPAR Vice-Chair D. Bilitza (USA), Executive Secretary, B. Reinisch (USA), Editor Members by Country: ARGENTINA: M. Mosert de Gonzalez, R. Ezquer AUSTRALIA: B. Ward, P. Wilkinson AUSTRIA: M. Friedrich BRAZIL: M. Abdu BULGARIA: I. Kutiev CHINA: Jiankui Shi, W. Wan, M.-L. Zhang FRANCE: D. Alcayde, R. Hanbaba CZECH REP: D. Buresova, L. Triskova, V. Truhlik GREECE: S. Kouris GERMANY: W. Singer, C. Stolle IVORY COAST: O. Obrou INDIA: K. Mahajan, S. Gupta, P.K. Bhuyan ITALY: S. Radicella, B. Zolesi JAPAN: K. Oyama, K. Igarashi, S. Watanabe NIGERIA: J. Adeniyi, E. Oyeyemi POLAND: I. Stanislawska, A. Krankowski, H. Rothkaehl R.O.C.: S.-Y. Su RUSSIA: A. Danilov, V. K. Depuev, T. Gulyaeva, G. Ivanov-Kholodny, A. Mikhailov, S. Pulinets, K.G. Ratovsky, I. Zakharenkova SOUTH AFRICA: A. Poole, L.-A. McKinnell SOUTH KOREA: K. Min SPAIN: D. Altadill UGANDA: J. B. Habarulema U.K.: P. Bradley, L.R. Cander, M. Rycroft ZAMBIA: Patrick Sibanda USA: D. Anderson, E. Araujo-Pradere, D. Bilitza, M. Codrescu, T. Fuller-Rowell, X. Huang, C. Mertens, B. Reinisch, L. Scherliess, J. Sojka, V. Wickwar, S-R. Zhang, I. Galkin Former Members: K. Bibl (USA, deceased), L. Bossy (Belgium, deceased) [IRI Chair from 1984 to 1992], L.H. Brace (USA, deceased), K. Champion (USA, retired), Y. Chasovitin (Russia, deceased), P. Dyson (Australia, retired), W. Hoegy (USA, retired), E. Kazimirovsky (Russia, retired), E. Kopp (Switzerland, retired), T. Maruyama (Japan), N. Matuura (Japan, retired), A.P Mitra (India, deceased), K. Rawer (Germany, retired) [IRI Chair from 1968 to 1984], K. Serafimov (Bulgaria, deceased)

Please cite this article in press as: Bilitza, D. The International Reference Ionosphere – Status 2013. J. Adv. Space Res. (2014), http://dx.doi.org/ 10.1016/j.asr.2014.07.032

D. Bilitza / Advances in Space Research xxx (2014) xxx–xxx

2. IRI-2012 The latest version of the IRI model, IRI-2012, includes several important improvements and new additions that lead to a more accurate representation of (1) the electron density and ion composition in the region from the F2 peak down to the E peak, (2) the solar cycle variation of the electron temperature, (3) the storm effects in the auroral E-region, and will for the first time include the representation of auroral oval boundaries and their magnetic storm induced movement to lower latitudes. These improvements and additions will be explained in the following sections. 2.1. New model for the bottomside electron density The electron density profile between the F1 ledge and the F2 peak is of special interest because of its effect on HF radio wave propagation and because of its non-negligible contribution to the ionospheric Total Electron Content (TEC). The profile in this region is described in IRI by the function   N e ðhÞ ¼ NmF 2  exp Z B1 =coshðZÞ; Z ¼ ðhmF 2  hÞ=B0

ð1Þ

which describes the dependence of electron density N e on height h. The profile is normalized to the F2 peak density NmF2 and the profile shape is described by the bottomside thickness parameter B0 and the shape parameter B1 . It is important to note that B0 is different from the often used half-density thickness parameter Y m that is defined as Y m ¼ hmF 2  h0:5 with the half-density height h0:5 defined as N e ðh0:5 Þ ¼ 0:5  NmF 2. In our case the height hx where B0 ¼ hmF 2  hx (or Z = 1) is the height where the density profile has in fact dropped to 0:24  NmF 2. IRI offers two options for the description of the global and temporal variations of the B0 and B1 parameters model and with IRI2012 is now adding a third one with the model of Altadill et al. (2009). All three models are based on ionosonde data. They vary in the volume and global coverage of the data used and in the mathematical formalism applied to represent these data. Gulyaeva (1987) used mainly data from mid-latitude stations and her model consists of a functional description of the observed correlation between hmF2 and h0:5 ; using Eq. (1) this can then be converted into a model for B0 . A much larger data base covering a wide range of global, seasonal, and solar cycle conditions was used by Bilitza et al. (2000) for their B0 and B1 models that grew out of an IRI Task Force Activity held at the International Centre of Theoretical Physics (ICTP) in Trieste, Italy in the 1990s. The B0 model consists of a table of representative values and an interpolation scheme for intermediate conditions. It describes variations with Local Time (LT), season, solar activity (R-12), and modified magnetic latitude (modip). The B1 parameter shows a marked change from

3

day to night and was modeled with the help of Epstein step-functions that vary from the daytime value of 1.9 to the nighttime value of 2.6 with smooth 1-h transitions at sunset and sunrise. Comparisons with newer ionosonde data from a wide range of stations found marked shortcomings of both model options in accurately representing the observed seasonal and solar activity variations of B0 and B1 (Lee and Reinisch, 2006; Lee et al., 2008;Chen et al., 2006; Blanch et al., 2007; Zhang et al., 2008; Sethi et al., 2009) and this was also confirmed with Millstone Hill incoherent scatter data by Lei et al. (2004). The IRI team therefore strongly encouraged the development of new B0 and B1 models that would overcome these shortcomings and take advantage of the much increased data base. Angling et al. (2009) took up the challenge and succeeded in developing a significantly improved model based on data from 27 globally-distributed ionosonde stations (DGS or DPS systems) for the years 1998–2006. The model uses a spherical harmonics formalism describing variations with modified dip latitude (modip), LT, month, and sunspot number. Evaluating their model versus the older IRI model options with their B0 data base, Altadill et al. (2009) find an improvement of up to 32% over the Bilitza et al. (2000) model and up to 40% over the Gulyaeva (1987) model and an improvement of up to 20% for B1. Because of these significant improvements the Altadill et al. (2009) option is the recommended default in IRI-2012. All three options are made available in IRI2012 to encourage continued evaluation with newer data sets. In Figs. 1(a and b) we compare the IRI predictions from these three options with observations of the Hainan ionosonde for the year 2004. We find that for this low latitude station all three B0 options have their shortcomings. Bilitza et al. (2000) is quite close to the monthly averages but shows the limitations of a step-type seasonal representation. Altadill et al. (2009) exhibits the correct annual variation but overestimates the observations. This is just one example that needs to be investigated more, but it shows the significant differences between the three B0 model options now available in IRI-2012. For B1 the new Altadill et al. (2009) model represents quite well the annual variation observed at Hainan while the two older models assume a constant value of 1.9. Improvements in the representation of the bottomside profile will also affect the ionospheric electron content (IEC) because the bottomside contributes about 15–40% to the total IEC. In addition the IRI-introduced B0 parameter has become an important parameter for characterizing the behavior of the bottomside ionosphere. Obrou et al. (2003), for example, find a close correlation of B0 from two stations near the magnetic equator with the strength of the equatorial electrojet. Lee (2011) pointed to the anomalous behavior of B0 during the recent extreme solar minimum, noting that, while the F2 peak density and height continue to decrease through 2008, B0 only decreased until 2007 and then started increasing again.

Please cite this article in press as: Bilitza, D. The International Reference Ionosphere – Status 2013. J. Adv. Space Res. (2014), http://dx.doi.org/ 10.1016/j.asr.2014.07.032

4

D. Bilitza / Advances in Space Research xxx (2014) xxx–xxx

Fig. 1. (a and b) Variation of the shape parameters B0 (top) and B1 (bottom) at noontime over the year 2004 as observed by the Hainan ionosonde (dots; the squares are the monthly averages) and as predicted by the Gulyaeva (1987) model (GUL: dashed curve), the Bilitza et al. (2000) model (TAB: dash-dot curve), and the Altadill et al. (2009) model (ABT: solid curve). The number of data points (NUM) is given in the lower right corner.

2.2. Inclusion of auroral boundaries At high-latitudes the influx of energetic solar wind and magnetospheric particles results not only in the beautiful display of the aurora borealis and australis, but can also have serious detrimental effects on technology in space. A demarcation of this region in IRI has therefore been a long sought after goal (Bilitza, 1995). The region is called the auroral oval (one in each hemisphere), typically ranging from 60° to 80° magnetic latitude and roughly centered at the magnetic poles. The oval is slightly elongated towards the night-side and has its largest extent at local midnight. During magnetic storms the oval expands and moves to lower latitudes. Different methods have been used to define and model auroral boundaries (for a review and references see Bilitza (1995)). A new model was recently presented by Zhang and Paxton (2008) based on global far ultraviolet (FUV) observations by the Global Ultraviolet Imager (GUVI) of the

Thermosphere Ionosphere Mesosphere Energetics and Dynamics (TIMED) satellite. The imager provides a much larger data base and better global and local time coverage than the in situ flux measurements on which the earlier models were based. GUVI provides cross-track scanned images of the Earth’s ultraviolet airglow and FUV auroral emissions. Radiances measured in the N2 Lyman–Birge– Hopfield (LBH) bands (LBHS: 140.0–150.0 nm and LBHL: 165.0–180.0 nm) can be used to obtain estimates of the mean energy (E0) and energy flux (Q) of precipitating electrons (Strickland et al. (1999)). Zhang and Paxton (2008) used the Atmospheric Ultraviolet Radiance Integrated Code (AURIC) of Strickland et al. (1999) and the Boltzman Three Constituent (B3C) of Daniell (1993) to construct tables that relate LBH radiances to the flux characteristics (E0, Q) of precipitating electrons. Based on GUVI data from 2002 to 2005 (44,000 images) they developed E0 and Q models describing variations with magnetic latitude and Magnetic Local Time (MLT) for

Please cite this article in press as: Bilitza, D. The International Reference Ionosphere – Status 2013. J. Adv. Space Res. (2014), http://dx.doi.org/ 10.1016/j.asr.2014.07.032

D. Bilitza / Advances in Space Research xxx (2014) xxx–xxx

different levels of magnetic activity. Fortuitously this time period included some of the most intense super-storms of the solar cycle and as a result the Zhang and Paxton (2008) model covers a much wider range of magnetic activity conditions (Kp = 0–9) than the earlier models. IRI-2012 uses the equatorward auroral boundary obtained from the Zhang and Paxton (2008) model at the threshold flux value of 0.25 ergs cm2 s1. Two adjustments had to be made before the model could be included in IRI. The first simply required converting the model from Kp dependence to ap dependence because IRI uses the 3-hourly ap index for the description of magnetic storm effects. This is trivial because the two scales are closely correlated. The second adjustment involves the coordinate systems used. The magnetic coordinates used by Zhang and Paxton (2008) are the Altitude Adjusted CGM coordinates (AACGM) of Baker and Wing (1989) and Bhavnani and Hein (1994) while IRI uses the Corrected Geo-Magnetic (CGM) coordinates of Gustafsson et al. (1992). Both coordinate systems use the International Geomagnetic Reference Field (Finlay et al., 2010) to trace from a point in space to the dipole geomagnetic equator and then trace back down along the dipole field line returning to the same altitude (CGM) or to the ground (AACGM) and use the so found geographic coordinates as the CGM or AACGM coordinates for the original point in space. CGM and AACGM coordinates are identical at the Earth surface but differences between the two increase with increasing altitude. The problem is easily resolved by using the Zhang–Paxton model in IRI with the CGM coordinates for altitude zero. Fig. 2 shows the variation of the CGM latitude with MLT for different levels of magnetic activity as given by IRI-2012. Including auroral boundaries in IRI is a first step towards a better representation of density and temperature features in IRI that are related to these boundaries such as the subauroral density trough and correlated temperature

5

peak. In Fig. 3 the IRI-predicted equatorward boundaries are shown for a Dynamics Explorer 2 orbit on a moderately disturbed day (July 13, 1982); the 3-hourly ap index for the shown time period was 144. The N e data shown are from the Langmuir Probe (LANG) and from the Retarding Potential Analyzer (RPA). The RPA measured also the electron temperature T e . We note the change from the relatively well behaved N e and T e variations equatorward of the boundary to the highly structured variations in the oval and polar cap that are not well represented by the IRI predictions that are also included in the figure. We can identify density troughs and temperature peaks near the IRI-predicted boundaries, however they are not always on the equatorward side of the boundary. But we have to keep in mind that IRI predicts the average boundary location which may differ from the one observed during a specific orbit. With the availability of near realtime or retrospective GUVI-type measurements, e.g., from TIMED/GUVI or DMSP/SSUSI, a more accurate specification of auroral boundaries can be achieved as was illustrated by Zhang et al. (2010). 2.3. Storm-time model for auroral E-region Increased particle precipitation during geomagnetic storms can produce significant electron density enhancements in the auroral E-region and lead to HF communication interruptions. A first attempt to describe these changes for IRI was made by McKinnell et al. (2004) and McKinnell and Friedrich (2007)with their Ionospheric Model for Auroral Zone (IMAZ). IMAZ uses a Neural Network (NN) formalism trained with EISCAT incoherent scatter radar measurements from 1984 and 1998 and data from radio wave propagation instruments on 50 rocket flights. Different influencing parameters were evaluated and best results were obtained with Magnetic Local Time, riometer absorption, local magnetic K index, F10.7 solar

Fig. 2. Variation of the IRI equatorward auroral boundary with Magnetic Local Time (MLT), Corrected Geomagnetic (CGM) latitude, and magnetic kp index.

Please cite this article in press as: Bilitza, D. The International Reference Ionosphere – Status 2013. J. Adv. Space Res. (2014), http://dx.doi.org/ 10.1016/j.asr.2014.07.032

6

D. Bilitza / Advances in Space Research xxx (2014) xxx–xxx

Fig. 3. Comparison of DE-2 CEP ( , red) and RPA ( blue) measurement with IRI (+) predictions during a moderately active period. The curves on top show the logarithm of the electron density and the curves at the bottom the electron temperature normalized to fit the left axis scale. The vertical lines indicate the time periods when the satellite crossed the IRI-predicted equatorward auroral boundary. Also shown is the satellite CGM latitude and MLT both normalized to the range 4–5 (e.g., CGM-lat = 90 corresponds to 4 and +90 to 5). The 3-hourly ap index for the time period is shown in the upper left corner. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

radio flux, and the neutral pressure. Neutral pressure provides a representation of the altitudinal and seasonal variations. The riometer absorption and K index track storm-related changes in the E-region. This modeling approach, however, depends on the availability of ground-based radio wave absorption (riometer) measurements and was difficult to seamlessly integrate with the existing quiet-time IRI model. It was provided as a separate, independent software code in IRI-2001 and IRI-2007. A different approach was used by Mertens et al. (2013a) and Mertens et al. (2013b). Following a similar methodology as Fuller-Rowell et al. (2000) had used for the F-region storm effects they determine the average storm-to-quiet time ratio of E-region electron density and its variation with magnetic activity. Their data sources are the TIMED SABER data. SABER is the Sounding of the Atmosphere using Broadband Emission Radiometry (SABER) instrument that measures limb radiances in several channels. Mertens et al. (2013a) and Mertens et al. (2013b) recognized that one of the channels, the 4.3 lm radiance measurement can be a good proxy for characterizing the nighttime E-region electron density in response to auroral precipitation. During daytime emissions at this wavelength are dominate by CO2 (v1, v2, v3) vibration–rotation bands and thus provide an excellent means for monitoring CO2 density measurement which was the prime reason for including the 4.3 lm channel. But during nighttime emissions from vibrationally excited NO+ become important and can be reliably separated from the background CO2 (v3) contribution. Mertens et al. (2013a) used precipitating electron energy flux measurements by the NOAA/POES satellite to prove that their SABER-deduced NO+(v) Volume Emission Rates (VER) are an excellent proxy for the incoming energy flux. Assuming charge neutrality and a

+ constant Oþ 2 to NO density ratio (because the reactions rates remain the same for quiet and storm conditions) they find

r ¼ VERStorm =VERQuiet  ½NOþ Storm =½NOþ Quiet  ½N e Storm =½N e Quiet

ð2aÞ

A power-law function is used to describe the dependence of the ratio r on the 3-hourly ap index rðap;km Þ ¼ C1  apC2 þ C3

ð2bÞ

The coefficients C1, C2, and C3 vary with magnetic latitude km as determined with the SABER-deduced NO+(v) VER data. Comparisons with nighttime and twilight incoherent scatter radar measurements from EISCAT and Sondrestrom during storm events show good agreement (Mertens et al., 2013b). One important shortcoming of this model is the limitation to nighttime and twilight hours. If this restriction was introduced into IRI we would get abrupt jumps in E-region electron density at dawn and dusk during storm periods. To avoid this unrealistic behavior we have as a preliminary first order approach assumed that the nighttime r(ap, km ) dependence can be also used for daytime. 2.4. Solar cycle dependent electron temperature model The IRI electron temperature model is based on the combination of global representations of T e at a number of fixed heights and the assumptions of constant gradients dT e /dh between these heights as illustrated in Fig. 4 (Bilitza et al., 1985; Bilitza, 1990). Epstein transition functions are used to get a continuous variation with height as shown by the solid curve in Fig. 4. The global models are based on

Please cite this article in press as: Bilitza, D. The International Reference Ionosphere – Status 2013. J. Adv. Space Res. (2014), http://dx.doi.org/ 10.1016/j.asr.2014.07.032

D. Bilitza / Advances in Space Research xxx (2014) xxx–xxx

7

Fig. 4. Build-up of the IRI electron temperature model combining global models at fixed heights and assuming constant gradient in between (as shown on the right). Epstein step functions are used to get a smooth altitude profile (solid curve). In this case the Bilitza (1985) option is shown with the global models of Brace and Theis (1981) at 300 km, 400 km, 1400 km, and 3000 km, the global model of Spenner and Plugge (1979) at 600 km, the CIRA model at 120 km and the model for the low-altitude peak height and peak temperature of Bilitza (1981) first introduced with IRI-79.

fitting spherical harmonics to the satellite data for the specific fixed height. IRI-2007 offers two different implementations of this modeling approach. Bilitza (1985) rely on the global models at 300 km, 400 km, 1400 km, and 3000 km that were developed with AE-C, ISIS-1, and ISIS-2 data by Brace and Theis (1981) and at 600 km based on the model developed by Spenner and Plugge (1979) with AEROS-A data (see Fig. 4). Truhlik et al. (2000) more recently utilized the global measurements of the Interkosmos 19, 24, and 25 satellites to develop global models at 350 km, 550 km, 650 km, 1400 km, and 2000 km with a more detailed description of the diurnal variation including the early morning overshoot that was not well represented in the Bilitza (1985) model. Additional fix-points for both model options are the lower boundary at 120 km and the height hem where T e reaches a local maximum (T em ). At 120 km and below thermal equilibrium is assumed resulting in T e ¼ T i ¼ T n and T n is determined by the COSPAR International Reference Atmosphere (CIRA, 1961). Like IRI, CIRA is being updated from time to time and the latest version is the NRLMSISE-00 model of Picone et al. (2002). With IRI-2012 this latest version is now used in IRI to determine T e at 120 km and below (the lower limit for IRI T e profiles is 60 km). In addition this model is also used to enforce the condition T e P T i P T n for all times and locations. Regarding the peak height hem and temperature T em their global variation was determined based on incoherent scatter data as described in Bilitza (1990). Comparisons with satellite and incoherent scatter radar measurements revealed some limitations of the Bilitza (1985) model in the representation of the observed seasonal and diurnal variations, e.g., the steep temperature maximum at sunrise (morning overshoot). The Truhlik et al.

(2000) model overcame many of these limitations with the help of a much larger data base. One of the still remaining shortcoming, however, is the fact that both models do not include variations with solar activity. Compared to latitudinal and diurnal changes the solar cycle variations are second order effects and of similar magnitude as the seasonal changes. Studies by Bilitza et al. (2007) and Truhlik et al. (2009) with a large data base of satellite in situ measurements have shown that T e can increase, decrease, or stay constant depending on the altitude, latitude, time of day, or season. This is because T e is determined by the balance of heating through photoelectrons that are created by the solar EUV irradiance, cooling through collisions with neutrals and ions, and heat conduction along magnetic field lines. All three terms increase with solar activity due to the increase in EUV flux, neutral density, neutral temperature, and electron and ion densities. Since the three terms compensate each other the net result can be a T e increase, decrease, or no change at all. Truhlik et al. (2012) have used the results of Bilitza et al. (2007) and Truhlik et al. (2009) to establish a solar activity correction term for the earlier Truhlik et al. (2000) model and this model is now included in IRI-2012. Fig. 5 shows two examples of the improvement reached with the new electron temperature model (Truhlik et al., 2012). In this figure several models are compared with satellite in situ data for a specific location (Millstone Hill), time of day (noon), season (equinox and summer, respectively) and altitude range (500–600 km and 760–940 km, respectively). The IRI-2007 T e model (dashed blue curve) does not vary with solar activity and therefor misrepresents the data by up to 500 K (20%). But we also note that overall the T e variation with solar activity is small for both

Please cite this article in press as: Bilitza, D. The International Reference Ionosphere – Status 2013. J. Adv. Space Res. (2014), http://dx.doi.org/ 10.1016/j.asr.2014.07.032

8

D. Bilitza / Advances in Space Research xxx (2014) xxx–xxx

Fig. 5. (a and b) Comparison of IRI-2012 electron temperature model with other models and in situ satellite measurements for noontime (LT = 11–13) at Millstone Hill for (a) equinox at 550 (±50) km and (b) summer solstice at 850 (±90) km. The temperatures are plotted vs the solar activity index PF10.7 = (F10.7D + F10.7_81)/2 with the daily and 81 day average of the solar radio flux at 10.7 cm (F10.7). The temperature data are shown as grey points. Also included are the mean (black dot) and standard deviations for the data and a least square fit to the data is shown as black line. The following models are included: the new IRI-2012 Te model (solid blue), the IRI-2007 Te model (dashed blue), the FLIP model (red), the Millstone Hill model (green) and the MSIS-86 neutral temperature (dashed orange).

examples. The new IRI-2012 T e model (solid blue curve) represents the observed solar activity variations quite well, even following the valley-shape observed in the 550 km case. Also included in these figures are the Field Line Interhemispheric Plasma (FLIP) physical model (Richards, 2001) (red curve) and the empirical model of Zhang et al. (2005) that represents many years of Millstone Hill incoherent scatter radar (ISR) measurements (green curve). FLIP is close to the data but exhibits a steady decrease in the low altitude case not following the observed valleytype shape. The Millstone Hill model is almost identical with the least-square fit line for the 550 km case, but underestimates the satellite data at 850 km by 500 K. This points to another problem we face when modeling the Te solar activity variations in the upper topside ionosphere. Discrepancies still remain between electron temperatures measured in situ by satellite and from the ground with incoherent scatter radars as for example shown in the comparisons of these two techniques at the DMSP altitude of 840 km altitude by Bilitza et al. (2007).

The first generation of IRI models for the ion composition was based on the pioneering work of Alexei Danilov and his group using a compilation of Russian rocket data including high-altitude rockets going up to topside altitudes (Danilov and Semenov, 1978; Danilov and Yaichnikov, 1985; Danilov and Smirnova, 1995). Their models give the percentage of O+, H+, N+, He+, NO+, Oþ 2 , and cluster ions as a function of solar zenith angle, latitude, season, and solar activity. These models give a first estimate of the distribution of ions in the ionosphere but are limited in their representation of latitudinal and diurnal variations because of the limited data base available for their development. Triskova et al. (2003) were able to produce a significantly improved model for the ion composition in the topside ionosphere by using carefully selected ion mass spectrometer measurements from the Interkosmos-24, AE-C, and AE-E satellites. The model takes advantage of the better global coverage provided by satellite measurements and uses the invdip latitude coordinate that is defined as invdip ¼

2.5. New model for the ion composition in the bottomside ionosphere IRI describes relative ion densities (ion composition) pj rather than absolute ion densities nj . This has the advantage that it simplifies P the needed enforcement of charge neutrality (N e ¼ j nj ) throughout the ionosphere. The absolute ion densities are obtained in IRI as nj ¼ pj  N e and thereby fulfill the charge neutrality equation. But it has the disadvantage that there are very few direct measurements of ion composition while the majority of measurements is of absolute ion densities, e.g., with ion mass spectrometers. The parameters measured by ground and space techniques are the absolute ion densities and the techniques do not always provide the total ion density that is required to calculate relative ion densities from the absolute ion density measurements. Ion composition modeling for IRI has long suffered from these data base limitations.

a invl þ b diplat aþb

where a = sin3 j diplat j and b = cos3(invl). Invdip is close to the dip latitude (diplat) near the magnetic equator and gets closer to the invariant latitude (invl) at higher latitudes and thus correlates well with the observed variation patterns of the topside ions (Truhlik et al., 2004). Compared to the older Danilov and Yaichnikov (1985) model the newer model also provides a more detailed description of variations over the year and uses the more appropriate Magnetic Local Time (MLT) to describe diurnal changes instead of the solar zenith angle that is more suited for the lower ionosphere. Noting discrepancies between the IRI model for the bottomside ion composition and AE-C ion mass spectrometer measurements, Richards et al. (2010) set out to develop a new model taking advantage of the fact that the photochemistry in the lower ionosphere is well established as many data-model comparisons have shown. Their approach uses

Please cite this article in press as: Bilitza, D. The International Reference Ionosphere – Status 2013. J. Adv. Space Res. (2014), http://dx.doi.org/ 10.1016/j.asr.2014.07.032

D. Bilitza / Advances in Space Research xxx (2014) xxx–xxx

the photochemistry that is applied in the FLIP model (Richards, 2001) and normalizes the total ion density with the IRI electron density. This approach is easy to implement in the region below 180 km where chemical equilibrium can be assumed. Higher up diffusion becomes more important and an iterative technique is applied to solve for the O+ density with the total ion density given by the IRI electron density. Their comparisons with AE-C ion mass spectrometer measurements show good agreement and also highlight the significant improvement the new model achieves over the older Danilov and Smirnova (1995) model. It also produces good agreement with the FLIP model itself, which obtains the ion densities by solving the continuity, momentum, and thermal equations. The new model will help to alleviate discrepancies noted by Nicolls et al. (2006) between nighttime airglow (630 nm and 557.7 nm) observations and IRI predictions and follows their suggestion of using straightforward ion chemistry for this part of IRI (Vlasov and Kelley, 2010). Model caveats are (1) that it does not account for auroral particle precipitation, and (2) also not for rapid convection. Partly this will be overcome by the normalization with a total ion density (=electron density) that includes these effects. 2.6. Improvement of the neutral atmosphere and magnetic field coordinates Being a COSPAR sponsored model, IRI needs to be compatible with the other COSPAR models in particular with the COSPAR International Reference Atmosphere (CIRA). To make IRI and CIRA compatible they have to fulfill two constraint conditions: (1) the IRI plasma temperatures should be in thermal equilibrium with the CIRA neutral temperature at 120 km and below and (2) they should not fall below the neutral temperature at any altitude. The IRI program therefore had to include the computation of the CIRA neutral temperature. With IRI-2012 the CIRA model has now been upgraded to the latest version which is the NRLMSIS-00 model (Picone et al., 2002). This is also of importance for the new model for the bottomside ion composition of Richards et al. (2010) that requires neutral densities for its photochemistry computations.

9

IRI-2010 makes use of the latest version of the International Geomagnetic Reference Field (Finlay et al., 2010) for its computation of magnetic coordinates using the latest sets of coefficients for 2010 and beyond to accurately represent the changes in the Earth’s magnetic field. For a better representation of the latitudinal changes of ionospheric parameters at high-latitudes IRI-2012 now also includes the Corrected Geomagnetic coordinates (CGM) of Gustafsson et al. (1992) that are needed, for example, for the representation of auroral boundaries. 2.7. Solar and magnetic indices Several solar and magnetic indices were added because of the requirements of some of the new model additions. IRI has traditionally used the 13-months running means of the sunspot number (R12) and of the ionosonde-based ionospheric-effective solar index IG (IG12). Fig. 6 shows these indices for the time period 2007–2017 as they were included in the indices file IG_RZ.DAT that is distributed with IRI. This file is updated at about quarterly intervals and Fig. 6 includes the most recent updates from June 2012, January 2013, September 2013, and December 2013. The R12 and IG12 predictions issued on June 2012 had been significantly scaled down by the time of the December 2013 release. It is important to be aware of these limitations when using IRI for times beyond the time of the last indices update; actually beyond 6 months before the last indices update, because we are using 13-month running means of these indices. In addition to IG12 and R12, IRI-2012 now also includes the F10.7 index which measures the solar radio flux at 10.7 cm wavelength and has been monitored from the ground near Ottawa, Canada since 1947 (Tapping, 2013). The index is included in IRI at a daily resolution, an 81-day resolution (3 solar rotations), and a yearly resolution to simplify its application in different parts of the IRI model. Fig. 7 shows the three indices for the time period 2010–2014. The daily index follows the small-scale structure related to the 27-day solar rotation, the 81-day index follows a half-year structure seen in the data, and the yearly index follows the 11-year solar cycle trend. While

Fig. 6. The 13-month running mean of sunspot number (left) and ionospheric index IG (right) for the time period 2007–2017. The different lines show the indices predictions included in the IG_RZ.DAT file as released at different times: June 2012, January 2013, September 2013, and December 2013.

Please cite this article in press as: Bilitza, D. The International Reference Ionosphere – Status 2013. J. Adv. Space Res. (2014), http://dx.doi.org/ 10.1016/j.asr.2014.07.032

10

D. Bilitza / Advances in Space Research xxx (2014) xxx–xxx

Fig. 7. Variation of the solar radio flux at 10.7 cm wavelength (F10.7) during the time period 2010–2014. In addition to the daily F10.7 indices (black) also shown are the 81-day average (light blue) and the 365-day average (dark blue). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

reviewing the daily F10.7 indices we found several dropouts (zero value) in the years 1959–1963 and have replaced these values with interpolations to avoid misrepresentations of IRI parameters for these cases. As explained in Section 2.2 IRI now also includes the magnetic index kp in addition to the already used ap index. 3. Ongoing activities towards future improvements of IRI Comparisons with IRI are often one of the first data evaluation tasks when a new data source becomes available. The goal is to compare the new ground or space data set with the prior empirical evidence accumulated in the IRI model. When consistent discrepancies are found and the reliability of the data is confirmed then efforts begin to improve IRI with this new data source. A number of IRI improvement efforts are currently underway we will focus here only on the two most important ones: a more accurate representation of the F-peak height hmF2 during quiet and storm times and the development of a Real-Time IRI. 3.1. Improved representation of hmF2 during quiet and storm times A strong focus of IRI activity has been on the global representation of the F2 peak height hmF2 because it is one of the most important parameter for radio wave propagation and because it is currently only indirectly modeled in IRI. The data-based representation of hmF2 in models relies almost exclusively on the CCIR (1966)-developed model for the propagation factor M(3000)F2 and the strong anticorrelation between hmF2 and M(3000)F2. M(3000)F2 is routinely scaled from ionograms and describes the

maximum usable frequency (MUF) that refracted in the ionosphere, can be received at a distance of 3000km (M(3000)F2 = MUF/foF2; foF2 is the F2 peak plasma frequency). To deduce hmF2 from M(3000)F2 the functional relationship has to account for the ionization below the F peak and different models have done this to different degrees. IRI uses one of the most advanced formulations that was developed using data from ionosondes as well as incoherent scatter radars and that depends not only on the E peak plasma frequency (foE) like most other models but also on the solar activity and the magnetic latitude (Bilitza et al., 1979). Shortcomings have been found to be mostly related to the limited time scales represented by the CCIRM(3000)F2 that, for example, smooth out the pre-reversal evening peak in hmF2 observed at low latitudes (Obrou et al., 2003). A recent comparison with South-African Digisonde measurements by Mbambo et al. (2013) showed that IRI predicted the diurnal seasonal structure at southern mid-latitudes quite well but in many cases overestimated the Digisonde-deduced hmF2 values by 10–30 km. IRI modelers have taken up the challenge and a number of new models have been developed for hmF2 and will be briefly reviewed here. Gulyaeva et al. (2008) used topside sounder data from the ISIS-1, ISIS-2, IK-19 and Cosmos-1809 satellites for the period of 1969–1987 to develop a model of hmF2 in terms of local time, season, geomagnetic latitude, geodetic longitude and solar radio flux; variations with these parameters are described by fitting polynomial functions to the hmF2 values that are deduced from the topside profiles assuming exponential profiles. Brunini et al. (2013) showed that the CCIR (1966) formalism of special spherical harmonics can be successfully applied directly to hmF2 instead of M(3000)F2. They

Please cite this article in press as: Bilitza, D. The International Reference Ionosphere – Status 2013. J. Adv. Space Res. (2014), http://dx.doi.org/ 10.1016/j.asr.2014.07.032

D. Bilitza / Advances in Space Research xxx (2014) xxx–xxx

conclude that hmF2 is best characterized by a Spherical Harmonics expansion of degree and order 15. Altadill et al. (2013) fitted spherical harmonics to a data base of Digisonde data from 26 worldwide distributed stations from the time period 1998–2006. The stations are well distributed in latitude but not in longitude. To remedy this limitation they introduce fictitious stations along lines of constant modified dip (modip) by assuming that differences in Local Time (LT) are equivalent to differences in longitude; this is somewhat similar to the screen points introduced in the CCIR (1966) analysis. Their final model consists of 9 sets of 610 coefficients (order 8 in modip, order 4 in longitude, and order 2 in day of year), one for each year. The new hmF2 model improves the fit to the Digisonde observations on average by 10% compared to the IRI prediction; the improvement can reach 30% at high and low latitudes. The model of Shubin et al. (2013) also uses spherical harmonics analysis but their data base consists of peak heights deduced from radio occultation data from CHAMP, GRACE, and COSMIC for low solar activity. Inputs for their model are geographical coordinates, month and the monthly F10.7 solar activity. Comparisons with Digisonde show standard deviations of 10–16 km for the new model while for IRI-2012 the deviations are slightly larger at 13–29 km. Average relative deviations are 3–4% for the new model and 9–12% for IRI. Maximal differences are found in the equatorial belt, over the oceans and in the polar caps. All of the models presented so far were dealing with the quiet-time hmF2. A number of studies have investigated storm effects on hmF2 and a few have also attempted to model these effects. Gulyaeva (2012) observes an anti-correlation between the storm-induced changes in hmF2 and NmF2 and model the seasonal, latitudinal, and solar cycle variation of the correlation coefficients with the help of topside sounder data from ISIS 1 and 2 and IK-19 and Cosmos1809. However, this model requires and dependents on an accurate representation of storm effects on the peak density NmF2. Blanch and Altadill (2012) model the storm-induced increase in hmF2 based on 32 storm events observed at a mid-latitude Digisonde station. They obtained good results by representing the observed double pulse in hmF2 by a combination of two Gaussian functions and with the coefficients of these functions changing with local time (of storm onset) and season. Good success (no false alarms) was achieved with using a drop of the IMF Bz component below 10 nT as trigger mechanism (in addition they found that they had to require a change larger than 20 nT in Bz within a time window of 3 h). Evaluations of these different hmF2 models are now underway and will lead to the inclusion of a much improved hmF2 model in IRI. 3.2. The Real-Time IRI An exciting new area of activity is the development of a Real-Time IRI based on the assimilation of real-time measurements into the IRI model. The standard IRI model

11

describes average conditions quite well and has shown excellent results in comparisons with other models (Shim et al., 2011; Shim et al., 2012). IRI performance can be even more improved if observations are available for a time of interest in real-time or retrospective. The simplest way is by way of an ionospheric-effective solar index that is obtained by adjusting IRI predictions to either ionosonde measurements (Bilitza et al., 1997), or GPS Total Electron Content (TEC) maps (Komjathy et al., 1998), or GPS slant TEC data (Hernandez-Pajares et al., 2002). The more complex assimilation of data into IRI has been used in a number of different approaches. Fridman et al. (2006) used the Tikhonov methodology with IRI and GPS data. The Electron Density Assimilative Model (EDAM) approach of Angling et al. (2009) is based on a weighted, damped least mean squares estimation (also referred to as Best Linear Unbiased Estimation (BLUE)) and assimilates mostly GPS data into IRI2007. Good results were obtained by Xinan et al. (2012) using a Kalman filter technique to assimilate GPS data, radio occultation data (CHAMP, GRACE, COSMIC, SAC-C, Metop-A, and TerraSAR-X), and Jason-1 and 2 altimeter TEC measurements into IRI-2007. Schmidt et al. (2008) represent the difference between GPS data and IRI globally and regionally with a multi-dimensional expansion in B-spline functions. Combing updating and assimilation techniques Pezzopane et al. (2011) first determine a ionospheric-effective sunspot number from comparing IRI to ionosonde F2 peak parameter (foF2, M(3000)F2) measurements and then after fully analyzing the ionograms applied a Kriging technique to assimilate the full electron density profile into IRI-2007. The IRI Real-Time Assimilative Mapping (IRTAM) of Galkin et al. (2012) is based on plasma frequency (foF2) measurements by the worldwide network of Digisonde stations (the Global Ionospheric Radio Observatory (GIRO)) and employs a linear optimization technique to obtain an improved global representation of foF2 for IRI every 15 min (http://giro.uml.edu/RTAM). Zhang et al. (2010) developed an algorithm for updating the IRI auroral boundaries (see Section 2.2) with Real-Time measurements by the GUVI instrument on the TIMED satellite and the SSUSI instrument on the DMSP satellite. The wide range of IRI Real-Time activities shows the great interest in the development of a data-assimilative capability with IRI. 4. Conclusions This paper is based on a status report given at the 2013 IRI Workshop in Olsztyn, Poland. It reviews the changes that were introduced with the release of the IRI-2012 version of the model and discusses two important ongoing activities, an improved representation of the F-peak height hmF2 and the development of a Real-Time IRI. IRI continues to evolve and IRI users will benefit from the many improvements introduced with IRI-2012: (a) A new model for the bottomside electron density resulting in a 32% improvement over IRI-2007.

Please cite this article in press as: Bilitza, D. The International Reference Ionosphere – Status 2013. J. Adv. Space Res. (2014), http://dx.doi.org/ 10.1016/j.asr.2014.07.032

12

D. Bilitza / Advances in Space Research xxx (2014) xxx–xxx

(b) For the first time inclusion of auroral boundaries and their movement with magnetic activity. (c) A representation of the storm-induced enhancement of electron density in the auroral E-region. (d) Inclusion of variations with solar activity in the IRI electron temperature model. A second-order effect but of importance because it changes direction depending on season, altitude, and time of day. (e) A new photo-chemistry-based model for the ion composition in the region below the F-peak. A significant improvement of the IRI-2007 model that was based on a limited amount of rocket and satellite data. (f) Upgraded versions of the models used internally to represent the thermosphere (NRLMSIS-00) and magnetic field (IGRF-11) and the addition of Corrected Geomagnetic (CGM) coordinates. The excellent performance and the wide usage of the IRI model are illustrated by the following highlights and usage statistics:  IRI is the standard for the ionosphere recommended by the European Cooperation for Space Standardization (ECSS), the International Standardization Organization (ISO), and the International Union of Radio Science (URSI).  IRI was one of the best performing models in the Electrodynamics Thermosphere Ionosphere (ETI) Challenge of the Coupling, Energetics, and Dynamics of Atmospheric Regions (CEDAR) program of the National Science Foundation (NSF) (Shim et al., 2011; Shim et al., 2012).  IRI is acknowledged in 8% of all 2012 Journal of Geophysical Research papers and 15% of all 2012 Radio Science papers.  Accesses to the IRIweb for computing and plotting IRI parameters shows roughly a 10-fold increase per year recently reaching the 1 million accesses per month mark.  IRI has been cited across a wide spectrum of journals including in 2012 the Journal of Geophysical Research, Geophysical Research Letters, Space Weather, Radio Science, Journal of Atmospheric and Solar-Terrestrial Physics, Advances in Space Research, Annales Geophysicae, Journal of Geodesy, Solar Physics, Journal of Asian Earth Science, Plasma Science and Technology, Space Science Review, Cosmic Research, Advances in Radio Science, Surveys in Geophysics, Planetary and Space Science, Chinese Journal of Aeronautica, Astrophysics and Space Science, Computer Physics Communications, Applied Optics, Computers & Geosciences, GPS Solutions, and Geochimica et Cosmochimica Acta.

Acknowledgments IRI is the result of modeling efforts by IRI Working Group members. All these contributions are essential for

the IRI success and are highly acknowledged by the author. I also want to acknowledge the many users of the model who have provided valuable feedback. A special thank you to Man-Lian Zhang who provided the Hainan ionosonde data used in Fig. 1. References Altadill, D., Torta, J.M., Blanch, E., 2009. Proposal of new models of the bottom-side B0 and B1 parameters for IRI. Adv. Space Res. 43, 1825– 1834. http://dx.doi.org/10.1016/j.asr.2008.08.014. Altadill, D., Magdaleno, S., Torta, J.M., Blanch, E., 2013. Global empirical models of the density peak height and of the equivalent scale height for quiet conditions. Adv. Space Res. 52, 1756–1769. http:// dx.doi.org/10.1016/j.asr.2012.11.018. Angling, M.J., Shaw, J., Shukla, A.K., Cannon, P.S., 2009. Development of an HF selection tool based on the electron density assimilative model near-real-time ionosphere. Radio Sci. 44, RS0A13. http:// dx.doi.org/10.1029/2008RS004022. Baker, K.B., Wing, S., 1989. A new magnetic coordinate system for conjugate studies at high latitudes. J. Geophys. Res. 94 (A7), 9139– 9143. http://dx.doi.org/10.1029/JA094iA07p09139. Benkova, N.P., Deminov, M.G., Karpachev, A.T., Kochenova, N.A., Kusnerevsky, Y.V., Migulin, V.V., Pulinets, S.A., Fligel, M.D., 1990. Longitudinal features shown by topside sounde data and their importance in ionospheric mapping. Adv. Space Res. 10 (#8), 57–66. Bhavnani, K.H., Hein, C.A., 1994. An improved algorithm for computing altitude dependent corrected geomagnetic coordinates, Scientific Report PL-TR-94-2310, Phillips Laboratory, 29 Randolph Road, Hanscom AFB, MA 01731-3010, USA. Bilitza, D., 1981. Models for ionospheric electron and ion temperature. In: International Reference Ionosphere – IRI 79, Report UAG-82, World Data Center A for Solar-Terrestrial Physics, NOAA, Boulder, Colorado, USA, pp. 11–16. Bilitza, D., 1985. Implementation of the new electron temperature model in IRI. Adv. Space Res. 5 (10), 117–122. http://dx.doi.org/10.1016/ 0273-1177(85)90193-0. Bilitza, D., 1986. International reference ionosphere: recent developments. Radio Sci. 21, 343–346. http://dx.doi.org/10.1029/RS021i003p00343. Bilitza, D., 1990. International reference ionosphere 1990, National Space Science Data Center, Report 90–22, Greenbelt Maryland, USA. Bilitza, D., 1995. Including auroral boundaries in the IRI model. Adv. Space Res. 16 (1), 13–16. http://dx.doi.org/10.1016/0273-1177(95)00093-T. Bilitza, D., 1997. International reference ionosphere – status 1995/96. Adv. Space Res. 20 (9), 1751–1754. Bilitza, D., 2001. International reference ionosphere 2000. Radio Sci. 36, 261–275. http://dx.doi.org/10.1029/2000RS002432. Bilitza, D., Reinisch, B.W., 2008. International reference ionosphere 2007: improvements and new parameters. Adv. Space Res. 42 (4), 599–609. http://dx.doi.org/10.1016/j.asr.2007.07.048. Bilitza, D., Sheik, N., Eyfrig, R., 1979. A global model for the height of the F2-peak using M3000 values from the CCIR numerical map. Telecommun. J. 46, 549–553. Bilitza, D., Brace, L.H., Theis, R.F., 1985. Modelling of ionospheric temperature profiles. Adv. Space Res. 5 (7), 53–58. Bilitza, D., Bhardwaj, S., Koblinsky, C., 1997. Improved IRI predictions for the GEOSAT time period. Adv. Space. Res. 20 (9), 1755–1760. Bilitza, D., Radicella, S., Reinisch, B., Adeniyi, J., Mosert, M., Zhang, S., Obrou, O., 2000. New B0 and B1 models for IRI. Adv. Space. Res. 25 (1), 89–95. Bilitza, D., Truhlik, V., Richards, P., Abe, T., Triskova, L., 2007. solar cycle variation of mid-latitude electron density and temperature: satellite measurements and model calculations. Adv. Space Res. 39 (5), 779–789. http://dx.doi.org/10.1016/j.asr.2006.11.022. Bilitza, Dieter, Brown, Steven A., Wang, Matthew Y., Souza, Jonas R., Roddy, Patrick A., 2012. Measurements and IRI model predictions

Please cite this article in press as: Bilitza, D. The International Reference Ionosphere – Status 2013. J. Adv. Space Res. (2014), http://dx.doi.org/ 10.1016/j.asr.2014.07.032

D. Bilitza / Advances in Space Research xxx (2014) xxx–xxx during the recent solar minimum. J. Atmos. Sol. Terr. Phys. 86, 99– 106. http://dx.doi.org/10.1016/j.jastp.2012.06.010. Blanch, E., Altadill, D., 2012. Midlatitude F region peak height changes in response to interplanetary magnetic field conditions and modeling results. J. Geophys. Res. 117, A12311. http://dx.doi.org/10.1029/ 2012JA018009. Blanch, E., Arrazola, D., Altadill, D., Buresova, D., Mosert, M., 2007. Improvement of IRI B0, B1 and D1 at mid-latitudes using MARP. Adv. Space Res. 39 (5), 701–710. http://dx.doi.org/10.1016/ j.asr.2006.08.007. Brace, L.H., Theis, R.F., 1981. Global empirical models of ionospheric electron temperature in the upper F-region and plasmasphere based on in situ measurements from atmosphere explorer C, ISIS 1 and ISIS 2 satellites. J. Atmos. Terr. Phys. 43, 1317–1343. Brunini, C., Conte, Juan F., Azpilicueta, Francisco, Bilitza, Dieter, 2013. A different method to determine the height of the F2 peak. Adv. Space Res. 51, 2322–2332. http://dx.doi.org/10.1016/j.asr.2013.01.027. CCIR, 1966. Comite´ Consultatif International des Radiocommunications, Report 340–1, Atlas of Ionospheric Characteristics, International Telecommunications Union, Gene`ve, Switzerland, ISBN 92-61-044174. Chen, H., Liu, L., Wan, W., Ning, B., Lei, J., 2006. A comparative study of the bottomside profile parameters over Wuhan with IRI-2001 for 1999–2004. Earth Planets Space 58, 601–605. CIRA, 1961. COSPAR International Reference Atmosphere, NorthHolland Publications, Amsterdam, The Netherlands. Daniell Jr., R.E., Modeling of optical signatures of electron spectra in the ionospheric heating experiments. In: Proceedings of the Seventh International Ionospheric Effects Symposium, SRI International, Arlington, Virginia, USA, 1993, p. 6B/5/1. Danilov, A.D., Semenov, V.K., 1978. Relative ion composition model at midlatitudes. J. Atmos. Terr. Phys. 40, 1093–1102. Danilov, A.D., Smirnova, N.V., 1995. Improving the 75 to 300 km ion composition model of the IRI. Adv. Space Res. 15 (2), 171–177. Danilov, A.D., Yaichnikov, A.P., 1985. A new model of the ion composition at 75 to 1000 km for IRI. Adv. Space Res. 5 (7), 75–79, and 107–108. de Larquier, S., Ruohoniemi, J.M., Baker, J.B.H., Ravindran Varrier, N., Lester, M., 2011. First observations of the midlatitude evening anomaly using super dual auroral radar network (SuperDARN) radars. J. Geophys. Res. 116, A10321. http://dx.doi.org/10.1029/ 2011JA016787. Finlay, C.C., Maus, S., Beggan, C.D., Bondar, T.N., Chambodut, A., 2010. International geomagnetic reference field: the eleventh generation. Geophys. J. Int. 183, 1216–1230. http://dx.doi.org/10.1111/ j.1365-246X.2010.04804.x. Fridman, S.V., Nickisch, L.J., Aiello, M., Hausman, M., 2006. Real-time reconstruction of the three dimensional ionosphere using data from a network of GPS receivers. Radio Sci. 41, RS5S12. http://dx.doi.org/ 10.1029/2005RS003341. Fuller-Rowell, T., Araujo-Pradere, E., Codrescu, M., 2000. An empirical ionospheric storm-time correction model. Adv. Space Res. 25 (1), 139– 148. Galkin, I.A., Reinisch, B.W., Huang, X., Bilitza, D., 2012. Assimilation of GIRO data into a real-time IRI. Radio Sci. 47, RS0L07. http:// dx.doi.org/10.1029/2011RS004952. Gulyaeva, T., 1987. Progress in ionospheric informatics based on electron density profile analysis of ionograms. Adv. Space Res. 7 (6), 39–48. Gulyaeva, T.L., 2012. Empirical model of ionospheric storm effects on the F2 layer peak height associated with changes of peak electron density. J. Geophys. Res. 117, A02302. http://dx.doi.org/10.1029/ 2011JA017158. Gulyaeva, T.L., Bradley, P.A., Stanislawska, I., Juchnikowski, G., 2008. Towards a new reference model of hmF2 for IRI. Adv. Space Res. 42, 666–672. http://dx.doi.org/10.1016/j.asr.2008.02.021. Gustafsson, G., Papitashvili, N.E., Papitashvili, V.O., 1992. A revised corrected geomagnetic coordinate system for epochs 1985 and 1990. J. Atmos. Terr. Phys. 54, 1609–1631.

13

Hernandez-Pajares, M., Juan, J., Sanz, J., Bilitza, D., 2002. Combining GPS measurements and IRI model values for space weather specification. Adv. Space Res. 29 (6), 949–958. Immel, T.J., Sagawa, E., England, S.L., Henderson, S.B., Hagan, M.E., Mende, S.B., Frey, H.U., Swenson, C.M., Paxton, L.J., 2006. Control of equatorial ionospheric morphology by atmospheric tides. Geophys. Res. Lett. 33, L15108. http://dx.doi.org/10.1029/2006GL026161. Komjathy, A., Langley, R., Bilitza, D., 1998. Ingesting GPS-derived TEC data into the international reference ionosphere for single frequency radar altimeter ionospheric delay corrections. Adv. Space Res. 22 (6), 793–802. Lee, C.-C., 2011. Equatorial B0 anomaly in September under extremely low solar activity. J. Geophys. Res. 116, A05325. http://dx.doi.org/ 10.1029/2010JA016394. Lee, C-C., Reinisch, B.W., 2006. Quiet-condition hmF2, NmF2, and B0 variations at Jicamarca and comparison with IRI-2001 during solar maximum. J. Atmos. Sol. Terr. Phys. 68, 2138–2146. http://dx.doi.org/ 10.1016/j.jastp.2006.07.007. Lee, C-C., Reinisch, B.W., Su, S.-Y., Chen, W.S., 2008. Quiet-time variation of F2-layer parameters at Jicamarca and comparison with IRI-2001 during solar minimum. J. Atmos. Sol. Terr. Phys. 70, 184– 192. http://dx.doi.org/10.1016/j.jastp.2007.10.008. Lee, I.T., Liu, J.Y., Lin, C.H., Oyama, K.-I., Chen, C.Y., Chen, C.H., 2012. Ionospheric plasma caves under the equatorial ionization anomaly. J. Geophys. Res. 117, A11309. http://dx.doi.org/10.1029/ 2012JA017868. Lei, J., Liu, L., Wan, W., Zhang, S.R., Holt, J.M., 2004. A statistical study of ionospheric profile parameters derived from millstone hill incoherent scatter radar measurements. Geophys. Res. Lett. 31, L14804. http://dx.doi.org/10.1029/2004GL020578. Lu¨hr, H., Xiong, C., 2010. The IRI2007model overestimates electron density during the 23/24 solar minimum. Geophys. Res. Lett. 37, L23101. http://dx.doi.org/10.1029/2010GL045430. Lu¨hr, H., Ha¨usler, K., Stolle, C., 2007. Longitudinal variation of F region electron density and thermospheric zonal wind caused by atmospheric tides. Geophys. Res. Lett. 34, L16102. http://dx.doi.org/10.1029/ 2007GL030639. Mbambo, M.C., McKinnell, L.-A., Habarulema, J.B., 2013. The variability and IRI2007-predictability of hmF2 over South Africa. Adv. Space Res. 52, 1798–1808. http://dx.doi.org/10.1016/j.asr.2013.06.022. McKinnell, L.A., Friedrich, M., 2007. A neural network-based ionospheric model for the auroral zone. J. Atmos. Sol. Terr. Phys. 69, 1459–1470. http://dx.doi.org/10.1016/j.jastp.2007.05.003. McKinnell, L.A., Friedrich, M., Steiner, R.J., 2004. A new approach to modeling the daytime lower ionosphere at auroral latitudes. Adv. Space Res. 34 (9), 1943–1948. http://dx.doi.org/10.1016/j.asr.2004.05.005. McNamara, L.F., Retterer, J.M., Baker, C.R., Bishop, G.J., Cooke, D.L., Roth, C.J., Welsh, J.A., 2010. Longitudinal structure in the CHAMP electron densities and their implications for global ionospheric modeling. Radio Sci. 45, RS2001. http://dx.doi.org/10.1029/ 2009RS004251. Mertens, C.J., Xiaojing, Xu, Bilitza, D., Mlynczak, M.G., Russell III, J.M., 2013a. Empirical STORM-E model: I. Theoretical and observational basis. Adv. Space Res. 51 (4), 554–574. http://dx.doi.org/ 10.1016/j.asr.2012.09.009. Mertens, C.J., Xiaojing, Xu, Bilitza, D., Mlynczak, M.G., Russell III, J.M., 2013b. Empirical STORM-E model: II. Geomagnetic corrections to nighttime ionospheric e-region electron densities. Adv. Space Res. 51 (4), 575–598. http://dx.doi.org/10.1016/j.asr.2012.09.014. Nicolls, M.J., Vlasov, M.N., Kelley, M.C., Shepherd, G.D., 2006. Discrepancy between the nighttime molecular ion composition given by the international reference ionosphere model and airglow measurements at low latitudes. J. Geophys. Res. 111, A03304. http:// dx.doi.org/10.1029/2005JA011216. Obrou, O.K., Radicella, S.M., Adeniyi, J.O., 2003. The equatorial electrojet and the profile parameters B0 and B1 around midday. J. Atmos. Sol. Terr. Phys. 65, 299–304. http://dx.doi.org/10.1016/S13646826(02)00336-X.

Please cite this article in press as: Bilitza, D. The International Reference Ionosphere – Status 2013. J. Adv. Space Res. (2014), http://dx.doi.org/ 10.1016/j.asr.2014.07.032

14

D. Bilitza / Advances in Space Research xxx (2014) xxx–xxx

Obrou, O., Bilitza, D., Adeniyi, J., Radicella, S., 2003. Equatorial F2-layer peak height and correlation with vertical ion drift and M(3000)F2. Adv. Space Res. 31 (#3), 513–520. Pezzopane, M., Pietrella, M., Pignatelli, A., Zolesi, B., Cander, L.R., 2011. Assimilation of autoscaled data and regional and local ionospheric models as input sources for realtime 3D international reference ionosphere modeling. Radio Sci. 46, RS5009. http://dx.doi.org/ 10.1029/2011RS004697. Picone, J.M., Hedin, A.E., Drob, D.P., Aikin, A.C., 2002. NRLMSISE-00 empirical model of the atmosphere: Statistical comparisons and scientific issues. J. Geophys. Res. 107 (A12), 1468. http://dx.doi.org/ 10.1029/2002JA009430. Rawer, K., Ramakrishnan, S., Bilitza, D., 1975. Preliminary reference profiles for electron and ion densities and temperatures proposed for the international reference ionosphere, Institut fu¨r physikalische Weltraumforschung, Scientific Report W.B. 2, Freiburg, Germany. Rawer, K., Bilitza, D., Ramakrishnan, S., 1978. International reference ionosphere, International Union of Radio Science (URSI), Brussels, Belgium. Rawer, K., Bilitza, D., Ramakrishnan, S., 1978. Goals and status of the international reference ionosphere. Rev. Geophys. 16, 177–181. Rawer, K., Lincoln, V., Conkright, R. (Eds.), 1981. International Reference Ionosphere – IRI 79, World Data Center A for SolarTerrestrial Physics, Report UAG-82. Boulder, Colorado, USA. Richards, P.G., 2001. Seasonal and solar cycle variations of the ionospheric peak electron density: comparison of measurement and models. J. Geophys. Res. 106, 12803–12819. http://dx.doi.org/10.1029/ 2000JA000365. Richards, P.G., Bilitza, D., Voglozin, D., 2010. Ion density calculator (IDC): a new efficient model of ionospheric ion densities. Radio Sci. 45, RS5007. http://dx.doi.org/10.1029/2009RS004332. Scherliess, L., Thompson, D.C., Schunk, R.W., 2008. Longitudinal variability of low-latitude total electron content: tidal influences. J. Geophys. Res. 113, A01311. http://dx.doi.org/10.1029/2007JA012480. Schmidt, M., Bilitza, D., Shum, C.K., Zeilhofer, C., 2008. Regional 4-D modeling of the ionospheric electron density. Adv. Space Res. 42 (4), 782–790. http://dx.doi.org/10.1016/j.asr.2007.02.050. Sethi, N.K., Dabas, R.S., Bhawre, P., Sarkar, S.K., 2009. Bottomside profile shape parameters during low solar activity and comparison with IRI-2007 model. J. Atmos. Sol. Terr. Phys. 71, 1935–1942. http:// dx.doi.org/10.1016/j.jastp.2009.08.003. Shim, J.S., Kuznetsova, M., Rasta¨tter, L., Hesse, M., Bilitza, D., Codrescu, M., Emery, B., Foster, B., Fuller-Rowell, T., Huba, J., Mannucci, A.J., Ridley, A., Scherliess, L., Schunk, R.W., Stephens, P., Thompson, D.C., Zhu, L., Anderson, D., Chau, J.L., Sojka, J.J., Rideout, B., 2011. CEDAR electrodynamics thermosphere ionosphere 1 (ETI) challenge for systematic assessment of ionosphere/ thermosphere models 1: NmF2, hmF2, and vertical drift using ground based observations. Space Weather 9, S12003. http://dx.doi.org/10.1029/ 2011SW000727. Shim, J.S., Kuznetsova, M., Rasta¨tter, L., Hesse, M., Bilitza, D., Butala, M., Codrescu, M., Emery, B.A., Foster, B., Fuller-Rowell, T.J., Huba, J., Mannucci, A.J., Pi, X., Ridley, A., Scherliess, L., Schunk, R.W., Sojka, J.J., Stephens, P., Thompson, D.C., Weimer, D., Zhu, L., Sutton, E., 2012. CEDAR electrodynamics thermosphere ionosphere (ETI) challenge for systematic assessment of ionosphere/thermosphere models: electron density, neutral density, NmF2, and hmF2 using

space based observations. Space Weather 10, S10004. http:// dx.doi.org/10.1029/2012SW000851. Shubin, V.N., Karpachev, A.T., Tsybulya, K.G., 2013. Global model of the F2 layer peak height for low solar activity based on GPS radiooccultation data. J. Atmos. Sol. Terr. Phys. 104 (2013), 106–115. http://dx.doi.org/10.1016/j.jastp.2013.08.024. Spenner, K., Plugge, R., 1979. Empirical model of global electron temperature distribution between 300 and 700 km based on data from AEROS-A. J. Geophys. 46, 43–56. Strickland, D.J., Bishop, J., Evans, J.S., Majeed, T., Shen, P.M., Cox, R.J., Link, R., Huffman, R.E., 1999. Atmospheric ultraviolet radiance integrated code (AURIC): theory, software architecture, inputs, and selected results. J. Quant. Spectrosc. Radiat. Transfer 62, 689. Tapping, K.F., 2013. The 10.7 cm solar radio flux (F10.7). Space Weather 11, 394–406. http://dx.doi.org/10.1002/swe.20064. Triskova, L., Truhlik, V., Smilauer, J., 2003. An empirical model of ion composition in the outer ionosphere. Adv. Space Res. 31 (3), 653–663. http://dx.doi.org/10.1016/S0273-1177(03)00040-1. Truhlik, V., Triskova, L., Smilauer, J., Afonin, V., 2000. Global empirical models of electron temperatures in the outer ionosphere for period of high solar activity based on data of three Intercosmos satellites. Adv. Space Res. 25 (1), 163–172. Truhlik, V., Triskova, L., Smilauer, J., 2004. New advances in empirical modeling of ion composition in the outer ionosphere. Adv. Space Res. 33 (6), 844–849. http://dx.doi.org/10.1016/j.asr.2003.06.006. Truhlik, V., Bilitza, D., Triskova, L., 2009. Latitudinal variation of the topside electron temperature at different levels of solar activity. Adv. Space Res. 44 (6), 693–700. http://dx.doi.org/10.1016/j.asr.2009.04.029. Truhlik, V., Bilitza, D., Triskova, L., 2012. A new global empirical model of the electron temperature with inclusion of the solar activity variations for IRI. Earth Planets Space 64 (6), 531–543. Vlasov, M.N., Kelley, M.C., 2010. Crucial discrepancy in the balance between extreme ultraviolet solar radiation and ion densities given by the international reference ionosphere model. J. Geophys. Res. 115, A08317. http://dx.doi.org/10.1029/2009JA015103. Xinan, Yue, Schreiner, William S., Kuo, Ying-Hwa, Hunt, Douglas C., Wang, Wenbin, Solomon, Stanley C., Burns, Alan G., Bilitza, Dieter, Liu, Jann-Yenq, Wan, Weixing, Wickert, Jens, 2012. Global 3-D ionospheric electron density reanalysis based on multi-source data assimilation. J. Geophys. Res. 117, A09325. http://dx.doi.org/10.1029/ 2012JA017968. Zhang, Y., Paxton, L.J., 2008. An empirical Kp-dependent global auroral model based on TIMED/GUVI data. J. Atmos. Sol. Terr. Phys. 70, 1231–1242. http://dx.doi.org/10.1016/j.jastp.2008.03.008. Zhang, S.-R., Holt, J.M., van Eyken, A.P., McCready, M., AmoryMazaudier, C., Fukao, S., Sulzer, M., 2005. Ionospheric local model and climatology from long-term databases of multiple incoherent scatter radars. Geophys. Res. Lett. 32, L20102. http://dx.doi.org/ 10.1029/2005GL023603. Zhang, M.-L., Wan, W., Liu, L., Shi, J.K., 2008. Variability of the behavior of the bottomside (B0, B1) parameters obtained from the ground-based ionograms at China’s low latitude station. Adv. Space Res. 42 (2), 695–702. http://dx.doi.org/10.1016/j.asr.2007.07.022. Zhang, Y., Paxton, L.J., Bilitza, D., 2010. Near real-time assimilation of auroral peak E-region density and equatorward boundary in IRI. Adv. Space Res. 46 (8), 1055–1063. http://dx.doi.org/10.1016/ j.asr.2010.06.029.

Please cite this article in press as: Bilitza, D. The International Reference Ionosphere – Status 2013. J. Adv. Space Res. (2014), http://dx.doi.org/ 10.1016/j.asr.2014.07.032