Aiming to understand the physical processes underneath the reversals events of geomagnetic field, different numerical models have been conceived. We considered here the so named ‘‘domino’’ model, an Ising–Heisenberg model of interacting magnetic macrospins aligned along a ring. This model was proposed by Mazaud and Laj (1989) and then applied by Mori et al. (2013) to study geomagnetic field reversals. The long series of the axial magnetic moment (dipolar moment or ‘‘magnetization’’) generated by the ‘‘domino’’ model are empirically studied by varying all model parameters. We present here some results which are slightly different from those given by Mori et al. (2013), and will provide our explanation on the presence of these differences. We also define the set of parameters that supply the longest mean time between reversals. Using this set of parameters, a large number of time series of axial magnetic moment are also generated. After de-noising the fluctuation of these time series and averaging them, we compared the resulting averaged series with the series of axial dipolar magnetic moment values supplied by CALS7k.2, and CALS10k.1b models, finding similar behavior for the all time series. In a similar way, we also compared the averaged 14,000 years long series of dipolar moment with the dipolar magnetic moment obtained by the model SHA.DIF.14k.

The differences between monthly mean values of the observed geomagnetic field and monthly values predicted by different models of the internal geomagnetic field (named “model biases”) for the time period 2000-2015 at several geomagnetic observatories are analyzed. We notice that increasing the maximum degree of the model is not always followed by the decrease of such “model bias”. The time series of these “model biases” reduced by their average resulted to be approximately the same for all models and should represent the external (non-modeled) contribution to the observed geomagnetic field. These time series for different observatories (close or away to each other) are compared and their power spectra are analyzed. Such spectra have common features like the annual and semi-annual variation with some possible sporadic cases of seasonal variation.

To provide insights on the paleosecular variation of the geomagnetic field and the mechanism of reversals, long time series of the dipolar magnetic moment are generated by two different stochastic models, known as the “domino” model and the inhomogeneous Lebovitz disk dynamo model, with initial values taken from the paleomagnetic data. The former model considers mutual interactions of N macrospins embedded in a uniformly rotating medium, where random forcing and dissipation act on each macrospin. With an appropriate set of the model’s parameter values, the series generated by this model have similar statistical behavior to the time series of the SHA.DIF.14K model. The latter model is an extension of the classical two-disk Rikitake model, considering N dynamo elements with appropriate interactions between them. We varied the parameters set of both models aiming at generating suitable time series with behavior similar to the long time series of recent secular variation (SV). Such series are then extended to the near future, obtaining reversals in both cases of models. The analysis of the time series generated by simulating both the models show that the reversals appear after a persistent period of low intensity geomagnetic field.

Theoretical estimates of the efficiency of the lower stratospheric ion-molecular reactions suggest an existence of another source of ozone (different from the well-known photo-dissociation mechanism). The conditions necessary for activation of autocatalytic ozone production require: (a) increased density of the low energy electrons (naturally produced at these levels by cosmic rays) and (b) reduced amount of water molecules above the tropopause. The short-lasting periods of a significantly decreased cosmic radiation—known as Forbush decreases (FDs)—provide a good opportunity for testing the above hypothesis. In this paper, we have compared the spatial distribution of three different FDs with changes in the lower stratospheric ozone density. We show that the strongest ozone depletion (beneath the ozone layer maximum) coincides very well with the largest magnitude of the FD. The time delay of ozone response to the applied forcing is too short, so the ozone depletion could not be attributed to the stratospheric circulations (as usually done). This result could be interpreted as an indirect confirmation of the importance of energetic particles, precipitating in the Earth's atmosphere, for the spatial-temporal variability of the lower stratospheric ozone density.

Using 12-year-long series of data (2001–2012) from geomagnetic observatories and repeat stations in Austria and its neighboring countries, a regional spatial–temporal (ST) model is developed based on the polynomial expansion consisting of latitude, longitude, and time of the geomagnetic field components and total magnetic field F. Additionally, we have used three different global models (CHAOS-5, POMME-9, and EMM2015), which are built on spherical harmonics up to a maximum degree Lmax and give the core field and crustal field separately. The normal field provided by the ST model and its “model bias”, which comprise the residuals of the differences between measured and predicted values, are calculated and the respective maps are shown. The residuals are considered an estimate of the local crustal field. In the case of global models, we have applied for each of these three methods to calculate the “model bias”: residuals of the differences between observed values and predicted values of the model, residuals of the differences between observed values and core field values of the model, and the average bias for the period 2001–2012. The normal field of the region of Austria provided by each global model is also calculated. Generally, the regional and global models yield relatively similar crustal  fields for the Austrian region, especially when the first method is used. The normal fields calculated by them are in good agreement with each other. Each of the global models directly provides the crustal field, and they are compared with the aeromagnetic data provided by aeromagnetic surveys over the Austrian region. The ST model is in better agreement with aeromagnetic data. We have also analyzed the secular variation over the region, which is calculated from the rate of change of normal field given by the ST and global models.

The static geomagnetic field of crustal origin is optionally calculated by the recent global geomagnetic field models. However, their description in global scale tends to miss some local characteristics. The same can be inferred for the rate of the geomagnetic field changes i.e. secular variation (SV). In order to depict some particularity of crustal field in local scale for a small region like Albania, two regional models are constructed: one based on the Legendre’s polynomials and the other based on a linear approximation. Both models use data from different measurement campaigns in the Albanian repeat station network and a few data from a neighbouring country. The residuals produced by these models and by the recent global models: EMM2015, POMME-9 and CHAOS-5 are calculated and compared. The SV from regional and global models are also calculated. Substantial differences between SV calculated by global models and regional models are observed.

The actual study is an attempt to analyze the influence of the main oxides on the behavior of heat cured fly ash/ slag blend geopolymer mortars activated with sodium hydroxide (NaOH). Fly ash and slag were decomposed into oxides and analyzed in oxide basis. Polynomial models for mortar flow workability and compressive strength were built based on 40 design points with different; oxide ratios, water content, NaOH levels and curing temperatures normalized appropriately. The same data were also modelled and analyzed in Design Expert package program Response Surface method using the historical data feature. The models show to be very stable as the error values are several orders of magnitude smaller compared to the respective coefficients. It was observed that main oxides such as CaO, MgO and SiO2 play an important role on the compressive strength of the mortars. On the other hand, both the methods used to build the models resulted in same equations, which indicates the consistency between the two approaches.

In this paper we employ a numerical approach to perform simulations of Maxwell distribution for several dimensionalities, based on the Central Limit Theorem. We show that by increasing the number of molecules of the gas N, the simulated distributions tend toward the respective theoretical distributions. Also, we observed that by increasing the model temperature n, the distribution shifted toward higher speeds, in agreement with theoretical results. The numerical simulations provide a physical definition of the concept of temperature. The codes used to perform the simulations are quite easy to construct and implement, while the results strikingly satisfy theoretical expectations. Furthermore, the actual approach makes it possible to skip the mathematical details and explain the distribution by just following the algorithm of simulations. We recommend such approach as a demonstrative tool that can be shown in a lecture class thus enriching the teaching quality and improving students’ understanding.

We have used a statistical model known as the “domino model” to simulate long time series of dipolar geomagnetic field. The simulated dipolar field time series, like the observed palaeomagnetic time series, has two stable states. The magnitude of the dipolar field in both cases oscillates irregularly between these states and these transitions, known as reversals of the dipolar geomagnetic field, seemingly occur at random. However there can be possible statistical indicators that serve as early signals about the incoming transitions. We applied some statistical tools like standard deviation, variance and power spectrum density (PSD) estimation to the simulated time series. All the statistical tools we have used show substantial change when approaching the eminent transition. Large values of the window width for the calculation of the standard deviation and variance indicate about the transition at earlier times. Also the PSD is affected when approaching the transition because the contribution to low frequencies is increased. Therefore we can identify these statistical tools as early indicators of future potential transitions.