Analysis Of The Influence Of DEM With Different Fineness On Precipitation Estimation Model

Topographic fitness is an important factor for precipitation estimation which is a basis for water resources development and natural disaster management.Therefore,the twodimensional empirical mode decomposition method(BEMD)was employed to the terrain fineness convermation of 1 km × 1 km spatial resolution DEM data in typical topography of humid area in humid area: Sichuan Basin,middle and lower reaches of the Yangtze River Plain,Wuyishan and Yungui Plateau.Then area rainfall models of multi-level fineness terrain were constructed to analyze the terrain fineness effect on precipitation of different terrain conditions.The main conclusions were as follows:(1)BEMD algorithm was applied to terrain fineness convermation,which was followed the principle of "keep the primary and remove the secondary" in terrain synthesis.The application results were 5-level terrain fineness,r1-r5,which were gradually decreasing in terrain fineness and surface roughness.The terrain smoothness of r5 is high,which meant that it is no neccessary to keep pushing down after r5.(2)terrain fineness effect of the precipitation related factors did exist: the variation of PPD in every study area was different.With the decrease of topographic fineness,the spatial "clutter" of PPD in Wuyishan was weakened first and then enhanced,while the spatial "clutter" of PPD in Sichuan Basin,the middle and lower reaches of the Yangtze River Plain and Sichuan Basin was gradually weakened.What's more,When the precipitation difference of leeward slope reached the maximum,the corresponding terrain fineness level was different.The plain area of Sichuan Basin and the middle and lower reaches of the Yangtze River were r2,the Wuyi Mountain was r4,and the Yunnan Guizhou Plateau was r3.(3)The simulation results of the 6-level terrain fineness area rainfall model including the original DEM showed that: if the terrain was too fine,the fine terrain data would increase the model noise;if the terrain was too rough,the correlation between the terrain parameters and precipitation in the model was weak,which would reduce the model accuracy.Therefore,it was necessary to select the most suitable terrain fineness level for the area rainfall modeling to achieve the highest estimation accuracy.(4)In different sub study areas,the terrain fineness level corresponding to the optimal precipitation estimation model was different: the best topographic fineness level of Sichuan Basin was r2,the plain area of the middle and lower reaches of the Yangtze River was r3,the Yunnan Guizhou Plateau area was r1,and the Wuyishan area was the original DEM.The correlation coefficients of the first three sub study areas were 0.94,0.95 and 0.91 respectively,which were 0.06,0.07 and 0.03 higher than those of the original DEM.