Filtering ionosphere parameters to detect trends linked to anthropogenic effects

Abstract

The great concern about the global warming observed in the troposphere has generated a large interest in the study of long-term trends in the ionosphere since the early 1990s, which has now become a significant topic in global change investigations. Some research works link ionosphere trends to anthropogenic sources such as the increase in greenhouse gas concentration, and others to natural causes such as solar and geomagnetic activity long-term changes, and secular variations in the Earth's main magnetic field. In all the cases, in order to analyze ionospheric trends, solar activity effect must be filtered out first since around 90% of ionosphere parameter variance is due to solar variations. The filtering process can generate 'spurious' trends in the filtered data series which may lead to erroneous conclusions. foF2 data series which include solar cycle 23 are analyzed in the present work in order to detect the effect of different filtering procedures on the determination of long-term trends. In particular, solar cycle 23 seems to have had an extreme ultraviolet (EUV) emission greater than that deduced from traditional solar EUV proxies during the maximum epoch and lower during the minimum epoch. When solar activity is filtered assessing the residuals of a linear regression between foF2 and Rz, or between foF2 and F10.7, this fact may bias trend values especially because it is at the end of the time series. The length of the period considered for trend assessment, the saturation and hysteresis effect of some ionosphere parameters, and the solar EUV proxy used are also considered in this study in order to quantify a possible spurious trend that may result as a by-product of a filtering process. Since trends expected as a consequence of anthropogenic effects are relatively small, these spurious effects may surely mask, or enhance, trends expected from anthropogenic origins.

Special issue “International CAWSES-II Symposium”

“…The F10.7 and Ap indexes were used to remove shorter variability of the parameter. Results of the analysis showed a long-term trend of −0.0075 MHz/year for the annual mean foF2, which was consistent with the results by Elias (2014). Bychkov et al (2014) investigated the correlation of 532-nm lidar returns at Kamchatka with the night ionospheric F2 layer and found the role of excited nitrogen ions in the formation of lidar signals.…”

foF2long‐term trend linked to Earth's magnetic field secular variation at a station under the northern crest of the equatorial ionization anomaly

“…In order to filter out solar activity effects, we took into account the period considered in the present analysis. As analyzed by de Haro Barbas and Elias [] and Elias [], solar cycle 19, for example, presents a strong maximum during which the ionosphere does not respond linearly to a further increase in EUV radiation. A kind of saturation takes place resulting in a breakdown of the linearity between f o F 2 and EUV with consequent persistent negative residuals during this period.…”

The Use of Monthly Mean Average for Investigating the Presence of Hysteresis and Long‐Term Trends in Ionospheric NmF2

“…Many insights have been gained based on this kind of methods on the topics of both hysteresis (e.g., Elias, 2014;Kane, 1992;Ozguc et al, 2008;Perna & Pezzopane, 2016) and long-term trends (e.g., Laštovička et al, 2016;Pham Thi Thu et al, 2016). In previous studies, yearly mean (or running) average methods are usually employed to smooth out the short-term, cyclic variations.…”

“…The existence of long-term trends will cause differences between the rising and declining solar cycle phase and thus contribute to the hysteresis (Ortiz de Adler & Elias, 2008). The hysteresis, on the other hand, implies changeable ionosphere-EUV relation on the solar cycle scale, which will affect the removing of solar influence by simple linear regression when extracting trends (Elias, 2014). Efforts have been taken to evaluate the ionospheric hysteresis and long-term trends from observations.…”

“…The resultant yearly values of ionospheric parameters and solar EUV flux are then used to construct a regression model to estimate the solar influence. Many insights have been gained based on this kind of methods on the topics of both hysteresis (e.g., Elias, 2014;Kane, 1992;Ozguc et al, 2008;Perna & Pezzopane, 2016) and long-term trends (e.g., Laštovička et al, 2016;Pham Thi Thu et al, 2016). Meanwhile, another method, that is, using monthly median/mean data to do the analysis for each month separately, was also found in literature (e.g., Bremer, 2008;Bremer et al, 2012).…”