Top concepts adopted by the current science on climate changes or atmospheric warming are not in agreement with the first principles of the physics. This paper presents a new understanding on the atmospheric behaviors. For example, the radiation is not the only factor that influences the air temperature, as the law of conservation of energy defines and as shown physically and mathematically in this article. The Sun is not the only heat source for the atmosphere because there is generation of heat at the Earth's surface by human activities. It is also shown that the water vapor is not a null effect and that the water vapor cannot be removed from the atmosphere for air temperature, greenhouse effect and climate changes considerations, in contrast to the current literature beliefs. The "feedback" concept is unfounded and invalid. The literature also says that "water vapor increases as the Earth's atmosphere warms", but this is also incorrect. The above equivocated understanding is accompanied by another one which believes that more water evaporates if the air temperature increases, but it is not in this way. These demonstrations and other authors' surveys showing that in the last decades the planet became wetter eliminate the literature concept that the water vapor does not have influence on the atmospheric warming/cooling. The conventional water cycle is related to the mass of water (mass of evaporation mass of precipitation) and then the physical and mathematical principles of the new hydrological cycle that includes the direct human influence are shown. The same is done for the carbon cycle. It is solved the problem on why the wind speed on Venus is very high above the cloud deck while it is stagnant below it, being this the same physical principle valid for the Earth's cloud cover. In the atmosphere, all the corresponding principles are the same, only their amounts change. It is demonstrated that the CO2 is not decisive for building and changing the temperatures of Venus, Mars, Mercury, Jupiter and Earth. Ice cores are not valid for "determining" "past" temperatures of the planet, because the mass of their air bubbles may be old, but the corresponding temperatures are not.