The 1-year（2009-2010） measurements are analyzed of the urban surface energy balance（SEB） obtained from the sensors located at three vertical layers of a 325-m tower in downtown Beijing.Results show that：（1） The measurements from the 325-m tower represent the SEB characteristics of the cities located in semi-humid warm-temperate continental monsoon climate zone.In a typical hot and rainy summer,cold and dry winter,the measured Bowen ratio is minimum in summer and maximum in winter.The Bowen ratio measured at 140 m for spring,summer,autumn,and winter are 2.86,0.82,1.17,and 4.16 respectively.（2） At the height of 140-m（in the constant flux layer）,the noontime albedo is ~0.10 for summer,~0.12 for spring and autumn,and ~0.14 for winter.The ratios of daytime sensible heat flux,latent heat flux,and storage heat flux to net radiation are 0.25,0.16,and 0.59 for clear-sky days,and 0.33,0.19,and 0.48 for cloudy days respectively.（3） Under clear-sky days,the nighttime sensible heat flux is almost zero,but the latent heat flux is greater than zero.For cloudy days,the nighttime sensible heat flux is slightly greater than the latent heat flux in winter.The nighttime upward heat flux is presumably due to the anthropogenic release（mainly latent heat for summer,while latent and sensible heat for winter）.
The numerical modeling of the impacts of urban buildings in mesoscale meteorological models has gradually improved in recent years. Correctly representing the latent heat flux from urban surfaces is a key issue in urban land-atmosphere coupling studies but is a common weakness in current urban canopy models. Using the surface energy balance data at a height of 140 m from a 325 m meteorological tower in Beijing, we conducted a 1-year continuous off-line simulation by using a coupled land surface model and a single-layer urban canopy model and found that this model has a relatively large systematic error for simulated latent heat flux. To improve the numerical method for modeling latent heat flux from urban surfaces, we combined observational analysis and urban land surface model to derive an oasis effect coefficient for urban green areas; to develop a temporal variation formula for water availability in urban impervious surfaces; and to specify a diurnal profile and the maximum values of anthropogenic latent heat release for four seasons. These results are directly incorporated into the urban land surface model to improve model performance. In addition, this method serves as a reference for studies in other urban areas.
