1. INTRODUCTION The Babcock-Leighton mechanism [Babcock, 1961; Leighton, 1969] outlines a basic picture of cyclic changes of the Sun's magnetic fields. First, the solar cycle starts from a poloidal field defined by the magnetic field confined in the polar areas of the Sun. Then, differential rotation converts this poloidal field into a toroidal configuration, giving rise to the emergence of active regions in the photosphere. As the solar cycle progresses, the magnetic field of active regions is dis-persed by turbulent convection and meridional flow. These transport mechanisms lead to the accumulation of magnetic flux of trailing polarity of decaying active regions at high solar latitudes, eventually reversing the polarity of the polar fields and building the next solar cycle. This concept led to the development of a distinct family of flux-transport numerical models that employ the Sun's differential rotation, supergranular diffusion and the meridional flows, to successfully represent many properties of observed long-term evolution of large-scale magnetic fields [DeVore et al., 1985; Wang et al., 1989]. The flux transport models were also extensively used to study effects of various parameters on the evolution of polar magnetic field and the solar cycle [Jiang et al., 2013]. For example, Baumann et al. [2004] have shown that the active region tilt described by Joy's law [Pevtsov et al., 2014], the diffusion and the rate of flux emergence have significant effect on the polar magnetic field. The speed of the meridional flow was found to affect the strength of solar cycle with slower meridional flow resulting in weaker solar cycles [Zhao et al., 2014]. Despite recent improvements, several questions remain open about flux transport models including the role of the not-well known variations in the speed of the meridional flow and scatter in orienta-tion of active regions (Joy's law). Both observations and numerical simulations indicate the importance of polar field as a predictor for strength of future solar cycle [e.g., Upton and Hatha-way, 2014]. On the other hand, the observational evi-dence of the importance of active region tilts on the strength of the solar cycle is inconclusive. The initial report by Dasi-Espuig et al. [2010] about finding a rela-tion between the mean active regions tilt of a given cy-cle and the strength of next cycle was questioned by Ivanov [2012]; McClintock, Norton [2013], and later, the results were revised by Dasi-Espuig et al. [2013]. Pevtsov et al. [2014] also argued that a relationship between the current surface activity (including active region tilt, flux emergence etc.) and the strength of the polar field may be complicated by a prior state of the polar field. For example, a strong surface activity may not necessary lead to a stronger polar field if it has a significant polar field of opposite polarity to cancel out. On the other hand, even a relatively modest surface activity may result in a strong polar field if the polar field of the previous cycle is weak. The question of the polar field strengths dependence on history is a long-standing issue in flux-transport research. For example, in their long-term flux-transport simulations Schrijver et al. [2002] found that the polar fields did not reverse during every cycle, e.g., a weak cycle would often fail to reverse strong polar fields. They suggested that this problem could be overcomed if the polar fields decayed away on timescales of 5-10 years. This idea of radial diffusion is not widely accepted now. Wang et al. [2002] showed that polar field reversals could be main-tained if the surface flow speeds were systematically higher in large-amplitude cycles than in weak ones. Whatever this or other mechanisms can explain the complicated relationships between succeeding activity cycles of different amplitudes and polar field strengths is still the subject of debate. The evolution of the polar magnetic field (and its re-versal) in the current solar cycle 24 has been recently studied by several researchers [Mordvinov, Yazev, 2014; Sun et al., 2015; Petrie, Ettinger, 2015; Tlatov et al., 2015]. Still, a complete understanding of the pro-cesses affecting the recent polar field reversal is missing. This justifies additional studies of the peculiarities of current solar cycle and its polar field reversals. In our study, we use a combination of synoptic observations and numerical modeling as described in detail in Sec-tions 2-5. Our findings are discussed in Section 6.