The structure of the boundary layer and the mechanism of heating over the Tibetan Plateau are examined using the data obtained from the First GARP (Global Atmospheric Research Program) Global Experiment and the Chinese Qinghai–Xizang (Tibet) Plateau Meteorological Experiment from May to August 1979. The meteorological elements near the plateau surface exhibit pronounced diurnal variations. There is a large ground–air temperature difference during the daytime generating a thin layer of superadiabatic lapse rates near the surface. The horizontal wind speed attains the minimum in the morning and the maximum in the evening. A deep and well-mixed layer of potential temperature is observed over the western and central plateau in the evening (1200 UTC). However, moisture is not well mixed vertically and water vapor mixing ratio is larger in the morning (0000 UTC) than in the evening. The mixed layer becomes shallower from the western to the central plateau and it disappears over the eastern plateau. There is a deep layer of large-scale ascent over the plateau. The upward motion and convective activity over the plateau are more intense in the evening than in the morning.
Before the onset of summer rains, sensible heat flux from the surface is the major source of heating on the plateau. The analysis of the ground–air temperature difference supports the hypothesis that dry convective elements rising from the heated surface provide the mechanism for tropospheric heating during this period. After the onset of the summer rains, however, the heat released by condensation is the primary source of heating over the eastern plateau. During the dry preonset period, the diabatic heating in the boundary layer is nearly balanced with the cold horizontal advection due to the prevailing westerly wind above the plateau. After the onset, the adiabatic cooling due to strong upward motions nearly balances with the heat released by condensation. The dry horizontal advection plays an important role in the moisture balance of the boundary layer.