Geological evidence suggests liquid water near the Earth's surface as early as 4.4 gigayears ago when the faint young Sun only radiated about 70 % of its modern power output. At this point, the Earth should have been a global snowball. An extreme atmospheric greenhouse effect, an initially more massive Sun, release of heat acquired during the accretion process of protoplanetary material, and radioactivity of the early Earth material have been proposed as alternative reservoirs or traps for heat. For now, the faint-young-sun paradox persists as one of the most important unsolved problems in our understanding of the origin of life on Earth. Here we use astrophysical models to explore the possibility that the new-born Moon, which formed about 69 million years (Myr) after the ignition of the Sun, generated extreme tidal friction - and therefore heat - in the Hadean and possibly the Archean Earth. We show that the Earth-Moon system has lost about 3e31 J, (99 % of its initial mechanical energy budget) as tidal heat. Tidal heating of roughly 10 W/m^2 through the surface on a time scale of 100 Myr could have accounted for a temperature increase of up to 5 degrees Celsius on the early Earth. This heating effect alone does not solve the faint-young-sun paradox but it could have played a key role in combination with other effects. Future studies of the interplay of tidal heating, the evolution of the solar power output, and the atmospheric (greenhouse) effects on the early Earth could help in solving the faint-young-sun paradox.

In this work, we study the probability density functions that best describe the $\gamma$-ray flux distribution of decade-long light curves of a sample of blazars. For averaged behavior over this period, there was maximum likelihood estimated log-stable distribution; for most sources leading to standard log-normal distribution ($\alpha=2$); however, other sources clearly displayed heavy tail distributions ($\alpha<2$), suggesting underlying multiplicative process of infinite variance. For sequences normalized using the found log-stable distributions, there was performed proposed novel non-stationarity and autocorrelation analysis. The former allowed to quantitatively evaluate non-stationarity of each source as maximizing log-likelihood rate of forgetting in modeled evolution of PDFs, also for evaluation of local variability allowing e.g. for anomaly detection suggesting changes of behavior. Discussed autocorrelation analysis looked at lag $l$ dependence of statistical behavior of all $\{(y_t,y_{t+l})\}$ points, described by various mixed moments - allowing to quantitatively point multiple characteristic time scales of the objects, for example, suggesting hidden periodic processes, with statistical interpretations of their contributions.