Since the industrial revolution, the greenhouse effect caused by greenhouse gas emissions has led to an increase in the net radiative forcing on the Earth’s surface, and the global mean surface temperature (GMST) has continued to rise, accompanied by multi-time scale oscillations including inter-annual decadal and multi-decadal oscillations. Changes in GMST, which reflect the integrated response of the climate system to the combination of external radiation forcing and internal natural variables, are one of the key indicators for understanding the impact of human activities on global climate change. Previous studies have basically determined the contribution of major physical modes of the global surface temperature, such as the El Nino-Southern Oscillation (ENSO), the Pacific Decadal Oscillation (PDO) and the Atlantic Multidecadal Oscillation (AMO) based on Empirical Orthogonal Function (EOF) method. In addition to the above climate modes, the physical processes affecting GMST changes also include volcanic eruptions, observation data system errors and so on. Based on the above research progress and these two basic assumptions (there is no long-term changing trend due to factors such as natural variabilities, volcanic eruption events, or system error of observation equipment; a single event does not affect the remotely related structure of the main physical mode), EOF decomposition is applied to the surface temperature of ERSST v3b and HadISST v1.1 ocean observation data in this paper, so as to obtain the TR warming trend item with ENSO, PDO and AMO removed, and thus, by using Ensemble Empirical Mode Decomposition (EEMD) method, to break down each TR time series into components on multiple different time scales and summarize them into low frequencies, high frequency and trend items. The land and ocean data (HadCRUT.4.6) are used to compare ocean data (ERSST v3b and HadISST v1.1) results and conduct a statistical analysis on the errors introduced by EMD methods. This study found that the main physical processes affecting changes in the GMST include global warming caused by the anthropogenic greenhouse gas emissions, the cooling effects caused by aerosols, the internal variables in the global climate system, the system errors caused by changes in the observing system, data noise, and other factors that have not yet been identified. Based on ERSST V3b and HadISST v1.1 observational data, this study also analyzed and determined the impact and contribution of each factor to GMST changes. This could be a key step to identify the effects of human activities and natural variability on global climate change. There are two main conclusive contributions in this paper. One contribution was to quantify the time scale and intensity of the impact of volcanic eruptions on GMST. We estimated the impact of five major volcanic eruptions since 1950 (Agung, 1963; Merapi, 1974; St. Helens, 1980; Pinatubo, 1991; Nabro, 2011). The cumulative effects of these five major volcano events in ERSST and HadISST approximately caused global warming to slow down by 0.32−0.47 and 0.45−0.67°C, respectively. Another major contribution is to find that the TR item of GMST still has a recent global warming mitigation phenomenon after the removal of the three major climate models, that is, there are some processes in the Earth’s climate system that are different from the above three physical models, which could significantly affect the TR changes. Therefore, global warming tends to slow down (warming hiatus). In modern climate conditions, anthropogenic radiative forcing continues to increase, and the heat absorbed by the global climate system continues to increase during global warming hiatus period. That is, even if the amount of heat energy used to heat up the surface of the global surface reduced, this “reduced” amount of heat still remains in the climate system, such as the mid-deep ocean. The results further underscore the necessity for understanding the critical physical processes that lead to global warming from the perspective of the energy balance of the climate system.