Yurdanur Tulunay

Forecasting of ionospheric critical frequency using neural networks

Multilayer perception type neural networks (NN) are employed for forecasting ionospheric critical frequency (foF2) one hour in advance. The nonlinear black-box modeling approach in system identification is used. The main contributions: 1. A flexible and easily accessible training database capable of handling extensive physical data is prepared, 2. Novel NN design and experimentation software is developed, 3. A training strategy is adopted in order to significantly enhance the generalization or extrapolation ability of NNs, 4. A method is developed for determining the relative significances (RS) of NN inputs in terms of mapping capability.

Variability of the interplanetary medium at 1 a.u. over 24 years: 1963–1986

Abstract A survey is presented of hourly averages of observations of the interplanetary medium, made by satellites close to the Earth (i.e. at 1 a.u.) in the years 1963–1986. This survey therefore covers two complete solar cycles (numbers 20 and 21). The distributions and solar-cycle variations of IMF field strength, B , and its northward component (in GSM coordinates), B z , and of the solar-wind density, n , speed, υ, and dynamic pressure, P , are discussed. Because of their importance to the terrestrial magnetosphere/ionosphere, particular attention is given to B z and P . The solar-cycle variation in the magnitude and variability of B z , previously reported for cycle 20, is also found for cycle 21. However, the solar-wind data show a number of differences between cycles 20 and 21. The average dynamic pressure is found to show a solar-cycle variation and a systematic increase over the period of the survey. The minimum of dynamic pressure at sunspot maximum is mainly due to reduced solar-wind densities in cycle 20, but lower solar-wind speed in cycle 21 is a more significant factor. The distribution of the duration of periods of stable polarity of the IMF B z component shows that the magnetosphere could achieve steady state for only a small fraction of the time and there is some evidence for a solar-cycle variation in this fraction. It is also found that the polarity changes in the IMF B z fall into two classes: one with an associated change in solar-wind dynamic pressure, the other without such a change. However, in only 20% of cases does the dynamic pressure change exceed 50%.

Characteristics of the mid-latitude trough as determined by the electron density experiment on Ariel III

Abstract Criteria are established to identify mid-latitude troughs by computer analysis of Ariel III electron density data and are used to study the variations of such troughs with time of day, season and magnetic activity. The statistics presented are based on over 1000 troughs and the features found are compared with previous work.

Temporal and spatial forecasting of ionospheric critical frequency using neural networks

The ionospheric critical frequency, ƒoF2, is forecast 1 hour in advance by using artificial neural networks. The value ƒoF2 at the time instant k of the day is designated by ƒ(k). The inputs used for the neural network are the time of day; the day of year; season information; past observations ƒoF2, the first difference Δ1 (k) = ƒ(k) = ƒ(k − 1); the second difference Δ2 (k) = Δ1 (k) − Δ1 (k − 1); the relative difference RΔ(k) = Δ1(k)/ƒ(k); geomagnetic indices Kp, ap, Dst, sunspot number, and solar 10.7-cm radio flux; and the solar wind magnetic field components By and Bz. This paper gives a new method, and it is the first application of neural networks for modeling both temporal and spatial dependencies. In order to understand the physical characteristics of the process and determine how important a particular input is, a test which shows the relative significance of inputs to the neural networks is performed at the output. The performance of a neural network is measured by considering errors. For the errors to be more meaningful, training and test times and times for comparison with other results are selected from the same solar activity period. Among the various neural network structures, the best configuration is found to be the one with one hidden layer with five hidden neurons, giving an absolute overall error of 5.88%, or 0.432 MHz.

Coincident observations of ionospheric troughs and the equatorial plasmapause

Abstract Explorer 45 traversed the plasmapause (determined approximately via the saturation of the d.c. electric field experiment) at near-equatorial latitudes on field lines which were crossed by Ariel 4 (~600km altitude) near dusk in May 1972 and on field lines which were crossed by Isis II (~1400km altitude) near midnight in December 1971 and January 1972. Many examples were found in which the field line through the near-equatorial plasmapause was traversed by Explorer 45 within one hour local time and one hour universal time of Ariel and Isis crossings of the same L coordinate. For the coincident passes near dusk, the RF electron density probe on Ariel detected electron density depletions near the plasmapause L coordinates when Ariel was in darkness. When the Ariel passes were in sunlight, however, electron depletions were not discernable near the plasmapause field line. On the selected near-midnight passes of Isis II, electron density depressions were typically detected (via the topside sounder) near the plasmapause L coordinate. The dusk Ariel electron density profiles are observed to reflect O + density variations. Even at the high altitude of Isis near midnight, O + is found to be the dominant ion in the trough region whereas H + is dominant at lower latitudes as is evident from the measured electron density scale heights. In neither local time sector was it possible to single out a distinctive topside ionosphere feature as an indicator of the plasmapause field line as identified near the equator. At both local times the equator-determined plasmapause L coordinate showed a tendency to lay equatorward of the trough minimum.

Variability of Turkish precipitation compared to El Niño events

Potential effects of the El Nino-Southern Oscillation (ENSO) warm extreme on the Turkish monthly precipitation totals between 1931 and 1990 were investigated using precipitation data from more than one hundred and eight Turkish meteorological stations. The data were normalized by using the annual total precipitation to minimize the probable topographical effects. A quiet standard annual variation was determined by month-by-month averaging for each annual month of the Neutral years. The anomalies in percent of the Turkish precipitation data were then calculated by subtracting El Nino-year monthly percent precipitation from the corresponding quiet standard value. The obtained anomalies were sorted after the El Nino high ENSO index month and analyzed by superposed epoch method. Distinct effects of high ENSO index months were found so that a large part of the month-to-month variability may be attributed to El Nino event. Spatially coherent and statistically significant precipitation responses to El Nino were shown in some regions of Turkey.

Temporal variations in the dawn and dusk midlatitude trough and plasmapause position

Abstract The temporal development of the latitudinal position of a 600 km midlatitude electron density trough at dawn and dusk during the period 25–27 May 1967, which encompassed a large magnetic storm, was measured by the RF capacitive probe on the polar orbiting Ariel 3 satellite. The substorm-related changes in the L coordinate of the trough minimum and the point of most rapid change of density gradient on the low latitude side of the trough are similar. Oscillations of the trough position at dusk are in phase with substorm activity whereas movement of the trough at dawn is only apparent with the onset of the large storm. Detailed model calculations of the plasmasphere dynamics assuming a spatially invariant equatorial convection E-field which varies in step with the Kp index produces a plasmapause motion which parallels the observed trough behaviour, particularly at dusk, and shows that the outer plasmasphere and possibly the trough region are characterized by complex fine structured variations due to the past history of the magnetosphere convection.

Temporal and spatial forecasting of the foF2 values up to twenty four hours in advance

Abstract Radio waves of a wide range of frequencies from very low frequency (VLF) to high frequency (HF), (broadly 3 to 30 MHz) can be propagated to great distances via the ionosphere. Since the largest variability occurs in the F-region the objective of this paper is to demonstrate a neural network model with the backpropagation algorithm which is designed to forecast the foF2 values of the highly nonlinear ionosphere up to 24 hours in advance. In other words, the model forecasts all values from 1 to 24 hours ahead. By using foF2 data for three European Ionospheric stations this neural network based model can forecast foF2 values both in time and in space for those three stations. The model seems promising for practical work since the root mean square errors involved are within reasonable limits.

An attempt to model the influence of the trough on HF communication by using neural networks

Trough is an interesting phenomenon in characterizing the behavior of the ionosphere, especially during disturbed conditions. The subject, which was introduced around the 1970s, is still attracting attention, especially during recent years. In HF communication, in particular, over the midlatitude ionospheric regions the electron density trough exhibits a phenomenon of abrupt gradients of electron densities in space and time which are directly reflected to foF2. Thus the performances of HF communications are directly affected. In this work an attempt has been made for the modeling to quantify the influence of the ionospheric midlatitude electron density trough on the ionospheric critical frequency foF2 by using neural networks. Data sets are used from the ground stations that include observations in the trough region. It has been demonstrated that the neural-net based approaches are promising in modeling of the ionospheric processes. Data generated by using statistical relationships are used to train the neural network. Then the trained neural network is used to forecast the ionospheric critical frequency, foF2, values 1 hour in advance for the cases when the probability of influence of the trough is high. Preliminary results will be presented to discuss the suitability of the neural-network-based approach in the modeling of complex processes such as the influence of the trough on foF2. The basic contributions of this work are 1) generation and organization of significant data for teaching complex processes, 2) neural-network-based modeling of a highly complex nonlinear process such as the influence of the trough on foF2 forecasting, and 3) general demonstration of learning capability by calculating cross correlations and general demonstration of reaching a proper operating point by calculating errors (that is, during the optimization process the neural network reaches the global minimum by using the gradient descent method).

Global electron density distributions from the Ariel 3 satellite at mid-latitudes during quiet magnetic periods

Abstract Electron density data returned by the Ariel 3 satellite have been separated into seasonal, diurnal, longitudinal and latitudinal groups. ‘Average’ electron density distributions for magnetic L values in the interval (1–11) are presented for each group and large scale features are summarised and discussed.