An analysis of data obtained from antarctic ice cores allows developing a simple theory that explains the physical processes of the past climate and quantitatively predicts the future climate. From the physical standpoint, there is not the slightest doubt about the reality of the greenhouse effect. The balance of energy falling on Earth from solar radiation and the microwave radiation leaving Earth into space indicates that the Earth's surface temperature is 32 K warmer than the equilibrium temperature that would be in the absence of the greenhouse effect. The theory of the greenhouse effect is also sufficiently well developed. The absorption spectra of the greenhouse gas molecules carbon dioxide .CO 2., methane .CH 4., and water vapor .H 2O. and their absorption band widening as pressure increases are well known, and the radiation convection theory, although quite complex, is nevertheless sufficiently developed for reliable computer modeling of Earth's atmosphere . The increasing combustion of fossil fuels (coal, oil, and gas) in the last century and a half has led to a substantial 40% increase in the main greenhouse gas .CO 2. in the atmosphere. The global warming resulting from this has been calculated by several world climate centers. Their results for the predicted temperature increase by the end of the 21st century over that at the beginning of the 20th century are in the range from 3 to 5.5 K. Such divergences arouse certain doubts in general society. One reason for such differences, the most obvious, is different model representations of the dynamics of further industrial development and, consequently, of future CO 2 emissions. On the other hand, there are also difficulties with the physical understanding of climate processes. For example, it is necessary to specify the cloud fraction in the model. Clouds cover approximately half Earth's surface on the average. This follows from the symmetry of convection: the vertical flows in cloudy cyclones and in cloudless anticyclones are equal. But we pose a qualitative question: Was the cloud fraction during glacial periods more or less than the contemporary cloud fraction? The factual answer is unknown. Current measurements, which correspond to a very narrow temperature range, give the sign of the trend inside the error limits . Meanwhile, the quantitative dependence is necessary for modeling the climate: the cloud fraction effect on the planet albedo is stronger than the effect of the glacial cover. The schoolhouse notion of the dew point is unfortunately inapplicable for describing cloudiness in a turbulent atmosphere . Another, more substantial difficulty for computer climate modeling is related to the need to predict the mass and heat exchanges between the upper layers of the ocean and its depths. The average temperature of the ocean surface is about 14 °C, and the ocean depths are substantially colder, 3 °C in all. Such a temperature distribution (seemingly contradicting thermodynamics) is maintained by the system of global thermohaline currents, called the Great Conveyor, which was discovered only in the 1980s as a result of research by Soviet and American oceanologists [4, 5]. The Great Conveyor begins with the sinking of cold saline waters near Greenland, flows in the depths of the Atlantic, curves around Antarctica, and surfaces near India and in the north Pacific Ocean. The time of the entire journey is about 1.5 ky. This means that the speed of the Great Conveyor current is two orders less than the speed of the ordinary wind-driven currents. Nevertheless, because of the large heat capacity of water, the influence of the Great Conveyor on climate is quite substantial.