Ionospheric Stations Research Articles

Ionosphere as a natural detector for investigations of solar EUV flux variations

It is proposed a method to determine solar EUV fluxes on the basis of data of ionospheric E-layer critical frequency measurements and of ionospheric activity index calculations. A new solar activity index was elaborated on the base of improved ionization theory in ionospheric E-region. It makes possible to carry out independent estimations of EUV fluxes and reciprocal calibration of EUV measurements, using information on ionospheric E-layer as data of the global natural detector of EUV emission. It is shown, that variations of the upper atmosphere temperature and density does not influence the instrumentation indications, and EUV flux responsible for E-layer ionization, can be determined due to E-layer critical frequency measurements. The new index may serve as a quantitative measure of EUV emission, at least, in lines 97.7 and 102.6 nm. Every mid-latitude ionospheric station may serve as a device for measurements of index – “equivalent flux” of EUV. Calibration error of this index is defined by a low error of radiophysical measurements of critical frequencies and is equal to ∼7%. The index allows also to investigate long-term variations and to expand a temporal range of research. The L α flux for almost four cycles of solar activity was calculated due to Slough and Chilton stations data. It is shown that variations of EUV fluxes approximately correspond to the main tendency of activity increase in the period from the 20-th to the 23-day cycle. Analysis of ionospheric index reveals an absence of noticeable long-term trends of EUV emission during 1957–2003 period. Use of this index allows to eliminate contradiction between EUV intensities observed onboard SME and UARS satellites, because fluxes in minima between the 21 and 22 cycles (SME epoch) and between the 22 and 23 cycles (a UARS epoch) are practically identical. In general all EUV cycles are similar to each other and it was not found a significant emission trend.

Seasonal variations of the ionospheric effects of geomagnetic storms at different latitudes of East Asia

We present the results derived from investigating the ionospheric effects of geomagnetic storm based on analyzing the data obtained at ionospheric stations located at different latitudes in the longitudinal sector of 90–130°E. There are considered geomagnetic storms of different intensities in the period from May, 2003 till March, 2004. During this period 15 storms are selected, which could be separated into different seasons. Our analysis of geomagnetic storms during different seasons suggests the following conclusions. (1) In summer during the initial positive phase of storms the positive disturbance in foF2 with the amplitude of up to 25% is observed at high latitudes. Then the disturbances are negative up to the end of the storm both at high and middle latitudes. The disturbance variations are similar. At low latitudes there are marked both positive and negative disturbances. (2) In winter the daytime disturbances are positive in the beginning of the storm at all stations under investigation. During the early recovery phase they are negative at high and positive at middle latitudes. The night disturbances are positive at high and negative at middle latitudes. At low latitudes the disturbances are positive both in daytime and at night during all storms. The disturbances change their sign near 30° geomagnetic latitude. (3) The equinox ionospheric variations during the storm are similar to those observed in summer: there are observed positive disturbances in the beginning of the storm and negative disturbances during the main and recovery phases both at high and middle latitudes. At low latitudes they are both positive and negative with high amplitudes. (4) In Hainan the intensive oscillations of foF2 are observed during the main phase of every storm.

Effectiveness of the IRI-2001-predicted N(h) profile updating with real-time measurements under intense geomagnetic storm conditions over Europe

It is well known that as long as variations in the ionosphere follow regular patterns, the latest version of the International Reference Ionosphere IRI-2001 model estimates sufficiently accurate NmF2 and other parameters relevant for the ionospheric effects on radio wave propagation. During geomagnetic storms, it is desirable that for the IRI-2001 model to be tested against available observations. This paper presents the comparison of the IRI-2001-generated electron density (N(h)) profiles with those updated using measured values over European area during October and November 2003 geomagnetic super storms. Ionospheric stations involved in this study are Athens, Chilton, Rome, Juliusruh and Tromso. These stations provide real-time ionospheric characteristics and N(h) profiles regularly within the framework of the EU COST271 Action on “Effects of the upper atmosphere on terrestrial and earth-space communications”. Comparative analysis shows that significant discrepancies do exist predominantly during the storm main phase. The model not always estimates correctly the phase and the magnitude of intense geomagnetic storm effects on the daytime F2 layer peak electron density at different European latitudes. Results are discussed in the context of real-time N(h) profile updating capabilities and effectiveness.

On spread-F in the ionosphere before earthquakes

Occurrence probability of the ionospheric spread-F in connection with earthquakes is analyzed. The F-layer is not close to the Earth (∼400 km), but in situ data could be obtained either by ionospheric sounders or by satellites. Data from the two Japanese ionospheric stations Kokubunji and Akita have been analyzed to find out long-term (a few weeks) variations of spread-F before and after earthquakes. Earthquakes with magnitudes M > 5 were taken into account. Only time intervals where geomagnetic variations are weak have been analyzed. It is shown that the probability of spread-F observations starts to decrease approximately 40 days before earthquakes, presents a minimum about 10 days before and then takes 1 month to recover the background level (therefore this increase lasts about 3 weeks after earthquakes). This effect exists if the distance between epicenters and the sounding station is less than 500 km.

On the global model for foF2 using neural networks

The use of neural networks (NNs) has been employed in this work to develop a global model of the ionospheric F2 region critical frequency, foF2. The main principle behind our approach has been to utilize parameters other than simple geographic coordinates, on which foF2 is known to depend, and to exploit the ability of NNs to establish and model this nonlinear relationship for predictive purposes. The foF2 data used in the training of the NNs were obtained from 59 ionospheric stations across the globe at various times from 1964 to 1986, on the basis of availability. To test the success of this approach, one NN (NN1) was trained without data from 13 stations, selected for their geographic remoteness, which could then be used to validate the predictions of the NN for those remote coordinates. These stations were subsequently included in our final NN (NN2). The input parameters consisted of day number (day of the year), universal time, solar activity, magnetic activity, geographic latitude, angle of meridian relative to subsolar point, magnetic dip angle, magnetic declination, and solar zenith angle. Comparisons between foF2 values determined using NNs and the International Reference Ionosphere (IRI) model (from Union Radio Scientifique Internationale (URSI) and International Radio Consultative Committee (CCIR) coefficients) with observed values are given with their root‐mean‐square (RMS) error differences for test stations. The results from NN2 are used to produce the global behavior of hourly values of foF2 and are compared with the IRI model using URSI and CCIR coefficients. The results obtained (i.e., RMS error differences), which compare favorably with the IRI models, justify this technique for global foF2 modeling.

On the global short‐term forecasting of the ionospheric critical frequency foF2 up to 5 hours in advance using neural networks

In this paper the use of neural networks (NNs) to forecast daily hourly values of the ionospheric F2 layer critical frequency, foF2, at any target geographic location up to 5 hours ahead is illustrated. The inputs used for the NN are universal time, day of the year, a 2 month running mean sunspot number (R2), a 2 day running mean of the 3 hour planetary magnetic ap index (A16), solar zenith angle, geographical latitude, magnetic dip angle, angle of magnetic declination, angle of meridian relative to subsolar point, and the four recent past observations of foF2 (F−3, F−2, F−1, and F0) from that target geographic location. The outputs of the NN are F+1, F+2,F+3, F+4, and F+5, representing the values of foF2 up to 5 hours ahead of F0. In this work, data from 40 worldwide ionospheric stations spanning the period 1964–1986, which include all periods of calm and disturbed magnetic activity, were used for training the NN. In order to test the predictive ability of the NN, the NN was verified with data from 10 stations not included in the training set that were selected for their remoteness from the trained stations. The results obtained from the NN are compared with the observed values of foF2 obtained from these selected verification stations. The performance of the NN is measured by calculating the root‐mean‐square (RMS) error difference between the NN model and measured station data. This paper illustrates that short‐term predictions of foF2 are much improved by including past observations of foF2 itself, in addition to those temporal and spatial inputs mentioned above, and that NNs can successfully be applied to the task of global forecasting.

DIAS Project: The establishment of a European digital upper atmosphere server

The main objective of DIAS (European Digital Upper Atmosphere Server) project is to develop a pan-European digital data collection on the state of the upper atmosphere, based on real-time information and historical data collections provided by most operating ionospheric stations in Europe. A DIAS system will distribute information required by various groups of users for the specification of upper atmospheric conditions over Europe suitable for nowcasting and forecasting purposes. The successful operation of the DIAS system will lead to the development of new European added-value products and services, to the effective use of observational data in operational applications and consequently to the expansion of the relevant European market.

Ionospheric perturbations linked to a very powerful seismic event

Ionospheric anomalies in association with the powerful Hachinohe earthquake ( M = 8.3 ) are derived using foF2 (F2 layer O-mode critical frequency) records from worldwide ionospheric stations. The results show that meaningful decreases in foF2 are identified only at several ionospheric stations near the epicenter, indicating that the anomalies extend to about 1500 km from the epicenter. These anomalies are observed a couple of days before and/or after the main shock and the duration of each anomaly is one-day. The anomalies occur predominantly during daytime.

Forecasting of Ionospheric Critical Frequency Using Neural Networks

由中国武汉电离层台站和澳大利亚Hobart台站的电离层F2层临界频率(f0F2)的资料,利用三层前向反馈神经网络(BP网络),提出一种提前24h预测f0F2的方法,该方法以前5天观测的f0F2数据拟合的5个系数以及太阳活动参数作为输入,以当天24 h的f0F2作为输出对网络进行训练,训练好的网络可以实现对f0F2提前24 h的预报.预测结果显示,利用神经网络预测的f0F2与实际观测结果变化趋势较一致,并且比IRI的计算结果更加准确.误差分析表明,在南半球Hobart(-42.9°,147.3°)台站比中国武汉站(30.4°,114.3°)的结果要好,在低年比高年要好,在冬夏季节比春秋季节稍好.本文说明利用神经网络对电离层参量进行预报是一种切实可行的方法.

Long-term trends in F2-layer parameters and their relation to other trends

The method developed and tested by the author earlier was applied to a detailed analysis of the h mF2 data obtained at the network of ionospheric stations to reveal long-term trends independent of the geomagnetic activity variations during the recent decades (nongeomagnetic trends). Unlike the results on foF2 published by the author earlier, the picture of h mF2 trends is not homogeneous. For 17 ionospheric stations positive significant trends are obtained. Six stations give negative trends in h mF2. For three stations no trends were derived. The comparison of the trends in f oF2 to the trends in the thermospheric density derived from satellite drag data shows their mutual agreement in the scope of current F-region theory. They also agree to the trends predicted in the 1990s for the CO 2 doubling. However, since the doubling has not yet happened, one has to look for other processes responsible for the trends. There is a possibility that the trends in thermospheric density and f oF2 may be of an anthropogenic origin. A hypothesis is considered according to which both these trends may be due to a decrease (negative trend) of the O concentration in the thermosphere. To relate this assumption to other trends the behavior of the E-layer parameters is considered. It is shown that positive trends in f oE are consistent with the negative trends of the [ NO + ] / [ O 2 + ] ratio, the latter being governed by the amount of NO. If there is a negative trend in [NO] in the E region during the recent decades, it is most probably due to the intensification (positive trend) of the eddy diffusion. This intensification should inevitably lead to a decrease of [O] in the entire atmospheric column above the E region. Such decrease may be responsible for the density depletion at satellite heights and f oF2 depletion in the F2-layer.

Ionospheric Effects of Geomagnetic Storms in Different Longitude Sectors

This paper analyzes the state of the ionosphere during two geomagnetic storms of a different intensity evolving in different sectors of local time in different seasons. There were used the data from a network of ionospheric stations located in the opposite longitudinal sectors of 80°-150° E and 250°-310° E.This analysis has permitted us to conclude that the detected differences in the variations of the disturbances are likely to be determined by the local time difference of the geomagnetic storm development, its intensity and by the different illumination conditions of the ionosphere.

Analysis of the Intense Magneticstorm of July, 2000 and of October, 2003 Using the Technique for Nowcasting of GPS TEC Data

利用武汉电离层观象台研制的GPS TEC的现报方法及现报系统,对东亚地区GPS台网的观测数据进行处理分析,特别对2000年7月14-18日和2003年10月28日至11月1日两次特大磁暴期间的数据进行了对比考察.文中分析了两次磁暴间的电离层响应,得到对应不同磁暴时段电离层TEC的不同变化情况,着重揭示了TEC赤道异常峰的压缩和移动以及赤道异常随时间的压缩一反弹-恢复的过程,并结合高纬电离层的部分响应机制进行了说明.结果显示,两次磁暴期的电离层响应表现出了各自不同的特点,从而反映出因季节变化引起的高纬电离层暴时能量注入的不同而造成的全球性电离层扰动的不同形态.由此看出,磁暴期间电离层TEC的变化直接与太阳扰动发生的时间及其对高纬电离层的耦合有关.若短时期内连续发生多次磁暴,则电离层反应更加复杂,不能简单地当做单一磁暴叠加处理.

Predicting the ionospheric F layer using neural networks

A new neural network (NN) based ionospheric model for the bottomside electron density profile over Grahamstown, South Africa (33.3°S, 26.5°E) has been developed and is referred to as the LAM model. This paper discusses the development of the F layer contribution to the LAM model. Archived data from the Grahamstown ionospheric station have been presented to various NNs, which have been trained to predict the parameters required to produce an electron density profile. Since the dataset was only made up of Grahamstown data, the model is currently a single station model. The input space designed for the F layer contribution to the LAM model consisted of various combinations of the following parameters: day number (DN), hour (HR), a measure of the solar activity, and a measure of the magnetic activity. The solar activity input was represented by a 2‐month running mean value of the sunspot number (R2), while the magnetic activity variable was represented by either a 24‐hour or 48‐hour running mean of the magnetic ak value (A8 or A16). The outputs from the NNs were the peak parameters and Chebyshev coefficients required to describe the shape and location of the F profile. This paper also discusses how NNs have been employed to provide an effective mechanism for determining the probability of the existence of an F1 layer. An F1 layer can exist in one of the following three states; F1 exists, no F1, or F1 exists in L condition state. A special NN was trained to provide the probability of F1 layer existence in each of these states. An L algorithm is applied to determine the shape and location of the profile under L condition state. In addition, a smoothing technique was designed to deal with discontinuities across the F1‐F2 boundary. It is shown that the NN‐based LAM model can successfully predict descriptions for the shape and location of the average profile for a given input set, and, in addition, that NNs can be employed to provide solutions to previously difficult prediction tasks such as the probability of F1 layer existence.

Mid-latitude ionospheric effects of a great geomagnetic storm

On March 13, 1989 magnetic storm effects on the mid- and low-latitude ionosphere were investigated. For this, peak electron density of F2-layer (NmF2) data from four chains of ionospheric stations located in the geographic longitude ranges 10°W–15°E, 55°E–85°E, 135°E–155°E and 200°E–255°E were used. Relative deviations of perturbed NmF2 from their respective quiet-time values were considered. Long-lasting negative storm effects were the dominant characteristic observed at middle latitudes, which occurred since the main phase of the storm. In general, the most significant negative disturbances were observed at middle-high latitudes. In the longitudinal sectors in which the storm started at day-time and pre-dusk hours, positive storm effects at middle and low latitudes were observed during the main phase. The role of some physical mechanisms to explain the ionospheric effects is also considered.

On ionospheric precursors of earthquakes in scales of 2–3 h

The probability of 2–3 h variation in electron concentration at different regions of ionosphere 1–3 days before earthquakes at distances R<500 km from vertical sounding stations was analyzed. Using 12 months of data sampled each 15 min at one ionospheric station it was revealed that before earthquakes with M>4.5 the number of variations with time scales of 2–3 h in the lower F-layer of the ionosphere f300 (real altitude of about 250 km; virtual altitude of 300 km) increases for 10 earthquakes, decreases for 4 and it does not changes for 2, on 16 cases we analyzed. Using the hourly data of 10 ionospheric stations collected for 20–30 years it was revealed that the number of disturbances in the foF2-frequency (real altitudes of about 300–350 km corresponding to the maximal electron density of F2 layer) with time scales of 2–3 h does not change before earthquakes with M⩽5.5 and slightly decreases (on 9%) before more strong earthquakes ( M>5.5). Possible models able to justify the phenomenology we observed are presented.

Day-to-day changes in experimental electron density profiles and their implications to IRI model

The electron density variability at fixed heights is studied for use in the International Reference Ionosphere IRI model. Monthly median, upper and lower quartile values were obtained for ƒ oF2, hmF2, B0 and B1 as deduced from ionograms. The IRI electron density profiles established with these values were then compared with the median and quartiles at fixed heights. Results are shown for the three ionospheric stations El Arenosillo (37.1 N, 353.3 E), Tucuman (26.9 S, 294.6 E) and San Juan (31.5 S, 290.4 E) as a function of solar activity, season and local time. As found by other authors the height of maximum variability, hvmax, is located below the peak electron density height ( hmF2) inall the cases. Values of differences between hmF2 and hvmax are analized. Variability defined as the interquartile difference and hvmax results are calculated from experimental electron density profiles measured at the three stations.

Predicted and measured bottomside F-region electron densityand variability of the D1 parameter under quiet and disturbed conditions over Europe

Modelling and forecasting of ionospheric parameters is very useful for different radio communication purposes. As long as variationsin the ionosphere form regular patterns, the empirical International Reference Ionosphere model, IRI 2000, provides sufficiently accurate corrections to the maximum electron density, N m F2, to predict the ionospheric effects on radio wave propagation. During geomagnetic storms, however, agreement between the IRI 2000 model and observations is still insufficient. This paper deals with the analysis of measured and model predicted F-region electron densities under quiet and disturbed conditions with the main emphasis placed on the distribution of the Fl-region daytime ionisation. Available electron density profiles obtained from ionograms for selected periods from several European ionospheric stations (Pruhonice, Ebro, Arenosillo) were compared with IRI 2000 model results. Comparative analysis shows that discrepancies do exist predominantly during the storm main phase. The model predicted daytime electron densities at the fixed F1-region heights are closer to observed values during summer than winter. Dependences of D 1 on solar activity and season are also analysed.

Regional mapping of F2 peak plasma frequency by spherical harmonic expansion

We use spherical harmonic analysis to obtain a 2-D map for the critical frequency of the F2 layer in two selected sectors; American and Europe⧹Africa within the area (15°S to 70°N latitude and 120°W to 55°E longitude). Basic data are monthly median f oF2 values obtained in the global network of ionospheric stations. Additionally, we compare our results for a few locations with those observed by INTERCOSMOS-19 topside sounder satellite and with the CCIR numerical maps, for different solar activity levels and seasons. It is shown that our monthly median regional maps are preferable to use of the world-wide CCIR numerical maps.

On the global behaviour of the day-to-day MUF variation

Using data from more than a hundred ionospheric stations spread worldwide, day-to-day variations with location and time of the conventional maximum usable frequency (MUF) decile values are investigated. New bounds of day-to-day MUF variability are calculated and compared with reference figures currently used by radio users. An hourly-daily validation process shows a clear improvement of spatial bias and model accuracy. A significant dependence on longitude is revealed that might be introduced into future specifications.

The large-scale isolated disturbances dynamics in the main peak of electronic concentration of ionosphere

The vertical sounding data at chains of ionosphere stations are used to obtain relative variations of electron concentration in the F2 ionosphere region. Specific isolated traveling large-scale irregularities are distinguished in the diurnal succession of the f cF2 relative variations records. The temporal shifts of the irregularities at the station chains determine their motion velocity (of the order of speed of sound) and spatial scale (of order of 3000– 5000 km , the trajectory length being up to 10000 km ). The motion trajectories of large-scale isolated irregularities which had preceded the earthquakes are reconstructed.