Abstract: Based on the multidimensional ensemble empirical mode decomposition (MEEMD) method, the long-term evolution of global land surface diurnal temperature range(DTR) trend from 1951 to 2019 was explored by using the Climate Research Unit (CRU) dataset. The results show that the diurnal variation of global surface temperature shows a downward trend, especially in the middle and high latitudes of the Northern Hemisphere (NH), and the spatial variability is considerable. The long-term evolution of global land maximum and minimum temperatures trends are mainly increasing, and the latter is more drastic. From the perspective of zonal averaged, five main negative bands of DTR are found. The decreasing trend gradually strengthens in NH, with accompanying cold tongues, while characteristic in the Southern Hemisphere shows strengthening first and then weakening.The instantaneous rate reveals the fact that the decrease rate of global DTR trends tends to slow down overall and turns positive in regions such as northern Australia, the Mediterranean coast, and southern South America in recent decades.