Both outputs from these two groups can well reproduce the spatial distribution characteristics of precipitation and air temperature in the simulated 30 years. As group RCM-CLM-CNDV is more sensitive to precipitation, the simulated LAI values in humid regions are slightly higher than observations. While with a larger spatial gradient, the LAI values in arid and semi-arid regions thus are appeared consistent with observations.
Part of Northwest-China and North-China (33°N-42°N, 100°E-120°E) were respectively isolated to discuss the relationship between drought events and the changes in vegetation. Several recorded drought events in the study area over the 30 years, including 1986-1987, 1994-1995, 1997-1998, and 2004-2005, can be significantly reflected in outputs from the two groups. Both the intensity and duration of these drought events in group RCM-CLM-CNDV are higher than in group RCM-CLM. A slower recovery of drought was perceived in group RCM-CLM-CNDV after comparing the changes of the Standard Precipitation Index (SPI). Also, changes in vegetation intensity, radiation balance, wind field, evapotranspiration, and water vapor flux in the outputs of the two groups of experiments were compared. Results suggested that the vegetation atrophy and the changes of surface conditions caused by a drought process in the semi-arid region in Northern China can increase the local surface albedo to a considerable extent, thus affecting the energy budget of the atmosphere. The vertical velocity anomaly in lower troposphere appears positive downward and can act as local sub-circulation to influence the weather system during the drought recovery stage. Reduced local evapotranspiration and hindered water vapor flux transmission can cause an unfavorable moisture condition. When the drought comes to its end phase, such resulting circulation signals can be transmitted to the upper layer and interfere with continuous precipitation processes, eventually prolonging the drought events.
Key words: North China, Drought Event, Land Surface Process, Vegetation Feedback, RegCM-CLM4.5