| Precipitation is an important climatic resource. It is closely related to flood monitoring and disaster management. Becaues China has a vast territory and complex topography, the distribution of monthly precipitation in China is uneven which brings huge difficulties in calculating areal precipitation on a fine-scale. Based on the previous researches, and considered the influence of atmospheric factors and topographic factors on the distribution of precipitation, a semi-theoretical and semi-empirical distributed model is established to estimate monthly precipitation in China over complex terrain using ground-based station data, final operational analysis data (NCEP-FNL) and digital elevation model (DEM). By this model, the spatial distribution of monthly precipitation during2000-2007is mapped at resolution of1km×1km. Through the comparison and analysis, it indicates that the model can reflect the difference in precipitation on different topographic orientation. It is able to qualitatively and quantitatively present the real distrbution of monthly precipitation in China. It has important value for disaster monitoring and ecological hydrology simulation.The following works are mainly completed in this paper:(1) Based on the theory of synoptic meteorology and NCEP-FNL data, started with physical mechanism of precipitation, the theoretical value of macro precipitation is calculated respectively by using water balance method and condensation function method.Through compared with ground-based observed data, the condensation function method is finally chosen to build the theoretical model of precipitation macroscopic field. The model can provide the spatial distributions of monthly precipitation macroscopic field quantity in china during the period of2000-2007with the resolution of1km*1km. The results accord with the objective law of the spatial and seasonal distribution of precipitation in china, such as precipitation in south is more than in north and in east is more than in west.(2) The concept of prevailing precipitation direction (PPD) which can be directly calculated by the monthly measurement data at meteorological stations is put forward to replace the wind direction in raining time in this paper. Then a multi-regression statistical model is established to estimate precipitation local field which is produced by terrain by using PPD, DEM, topographic orientation, etc. Through comparison with the prevailing wind derived from FNL and validation by monograph of regional climate, it shows that PPD can reflect the regional precipitation climatic characteristics. Then using the sign of cosine value of angle between PPD and macro-topographic orientation, the distributions of windward slopes and leeward slopes can be identified quantitatively.(3) Combine precipitation macroscopic field and precipitation local field, a semi-theoretical and semi-empirical model is established to estimate the distribution of monthly precipitation in China over the complex terrain by optimized stepwise regression analysis. By the test of verification stations, mean absolute errors are2mm and17.4mm for January and July respectively, and mean relative errors are21.8%and12.9%for each.(4) The model which the paper puts forward is named distributed model. It is compared with other three models whose are the general regression model, Guan model and TRMM model. On the one hand, through the comparison of errors, it shows that the simulation accuracy of distributed model is better than the other three models. On the other hand, by applying four models in Tianshan Moutain and Wuyi Mountain, it shows that unlike other three models, the distribution of estimated precipitation devired from distributed model is reasonable that can reflect difference of precipitation between the windward slope and the leeward slope. |
PDF Full Text Request