The seismic performance of a caisson structure under two types of models with a saturated sandy foundation (CSS) and an expanded polystyrene (EPS) composite soil foundation (CES) are studied using shaking table tests. The macro phenomena of the two different foundation models are described and analyzed. The effects of the replacement of EPS composite soil on seismic-induced liquefaction of backfill and the dynamic performance of a caisson structure are evaluated in detail. The results show that the excess pore water pressure generation in the CES is significantly slower than that in the CSS during the shaking. The dynamic earth pressure acting on the caisson has a triangular shape. The response of horizontal acceleration, displacement, settlement, and rotation angle of the caisson in the CES is smaller than that in the CSS, which means the caisson in the CES has a better seismic performance. Furthermore, the out-of-phase phenomenon between dynamic earth thrust and inertial force in the CES is more obvious than that in the CSS, which is beneficial to reduce the lateral force and improve the stability of the caisson structure.

The surface energy balance is a key issue in land surface process research and important for studies of climate and hydrology. In this paper, the surface energy fluxes (net radiation, ground heat flux, sensible heat flux and latent heat flux) at the Tanggula (TGL) and Xidatan (XDT) sites were measured and the distributions of the regional surface energy fluxes on the Tibetan Plateau were obtained using a revised surface energy balance system (SEBS) model. The results show that the surface energy fluxes have obvious seasonal variations. At both sites, the sensible heat flux is highest in spring and lowest in summer, and the latent heat flux is highest in summer and lowest in winter. The high elevation, snow cover, freeze-thaw process, precipitation, vegetation and soil texture are important influencing factors for land surface energy fluxes. The time-phase difference between the net radiation and ground heat flux for bare soils is estimated to be 2-3 hr. The ratio of ground heat flux and net radiation ranged from approximately 0.18 to 0.33, and a parameterization scheme for the remote sensing of ground heat flux over the Tibetan Plateau bare soil in summer is developed. The simulation results of the regional surface energy fluxes show that the distributions of surface parameters, such as vegetation, soil texture and soil moisture content, are important for understanding regional changes in the surface energy fluxes.