Solar radiation provides the primary energy resources to the earth, which distribution is the basic power of the physical processes in nature, such as snow/ice melting, photosynthesis and evaporate-transpiration. Thus, it is extremely necessary to explore the solar radiation variation in the researches on hydrothermal regime, land surface geographical process and the use of solar energy. With the rugged terrain in the mountain regions on basis of current conditions, the terrain factors of slope, aspect and hill shade by surrounding terrain varies, and the received solar energy on the land surface appears diversity at different conditions, which causes the complex spatial distribution of solar radiation. Taking the influence of topographical factor on the solar radiation distribution into consideration, the distribution solar radiation models based on terrain analysis were built to simulate the total solar radiation over rugged terrains since 1980s. Latter，the impact of surrounding terrain and weather condition on the received solar radiation considered, several methods such as amending sunshine hours, introducing transform indexes, sky view factorand topographic correction index, were used to improve the simulated values of distributed solar radiation models. Nevertheless, the calculated values of hillshade in the current researches exists the certain bias, dueling to the spatial scale of DEM and the accuracy of the measured meteorological data, and the accuracy and temporal resolution of solar radiation model is needed to improved. A new hourly distributed solar radiation model over mountains is built in this paper, making use of ASTER GDEM data to amend the current Yang-Kun solar radiation model at single point under clear weather conditions. The total solar radiation is divided into the direct solar radiation and diffuse solar radiation. Shade-sunshine judgment factor is introduced into the calculation of direct solar radiation, as well as topographic correction parameter, with aim to quantify the hillshade effects of slope itself and its surrounding terrains. Meanwhile, the parameter of sky view factor is also introduced to simulate the values of possible diffuse solar radiation which comes from nearby eight directions. Then the meteorological parameters of atmospheric pressure, air temperature, air relative humidity and sunshine hours provided by two automatic weather stations (AWS) named Daxigou and Kongbingdou, as well as ASTER GDEM, are collected as the input parameters of the model, and the model is used to simulate hourly instantaneous solar radiation with the resolution of 30 meters at the head of Ürümqi river basin in Xinjiang province. Lastly, the simulated values of solar radiation at AWS positions are validated against the measurements of solar radiation to assess the applicability of the model. The results in this research can be shown as following: 1) the model built has the high accuracy, and the determination coefficients (R2) between the simulated and measured values of solar radiation at Daxigou and Kongbingdou weather stations are 0.94 and 0.83, respectively; 2) for the lack of the measured sunshine hours at Kongbingdou station, the accuracy of solar radiation simulation at this station appears less than that at Daxigou station in cloudy day, which implies that this model is highly depend on the sunshine hours; 3) the model proves the perfect accuracy of simulation in the clear day except for the sunrise and sunset hours, with the determination coefficients(R2) between the simulated values and the measurements at Daxigou and Kongbingdou stations are 0.96 and 0.89, the mean relative error of 2.6% and 0.8%, respectively. While the model performs the poor simulated values of solar radiation in the cloudy day, with the determination coefficients (R2) between the simulations and the measurements at Daxigou and Kongbingdou stations are only 0.43 and 0.56, the mean relative error of 34.5% and 25.9%, respectively, which implies that the simulated values of the model built are strongly depend on solar radiation in clear day and topographic correction indexes. Generally speaking, the distributed solar radiation model built in this paper, considering the influence of topographic factors on the received solar energy over rugged terrains, proves the perfect simulation of solar radiation, especially under the clear conditions. This implies that the quantification of terrain characteristic will be helpful to improve the accuracy the simulation of the physical parameters near the land surface.