| The ability of simulating the atmosphere of high resolution numerical weather prediction model has been a hot topic within the meteorology research field. Based on the analysis of high resolution numerical model's ability in atmospheric multi-scale movement simulation, a new point of view is presented on this issue. The atmospheric boundary layer(PBL) observed data which from July 2001 uses in the article comes from state project 973's research experiment--The environmental evolution of existence and Trends of Aridification in the future for northern China. Detailed analysis of PBL atmospheric multi-scale movement characters of this region will be presented first. Then ARPS (Advanced Regional Prediction System) mesoscale numerical model will be used in a nine nested numerical simulation experiment which uses NCEP analysis data for the initial and boundary conditions. The simulation results of u, v, w,θ, p for different model resolutions will be analyzed with spectrum analysis and wavelet analysis. Finally, WRF (Advanced Research Weather Research and Forecasting Model) mesoscale numerical model will be used in three four nested numerical simulation experiments, which use NCEP analysis data for the initial and boundary conditions, with different PBL parameterization schemes in the same region. The simulation results of ten meters tall wind u; v and two meters tall temperature for different model resolutions and PBL parameterization schemes will be analyzed with spectrum analysis and wavelet analysis.The studies show that, there are multi-scale features in real atmosphere and it has much difference between various physical quantities' multi-scale features. The physical quantity such as temperature which is affected by outer cyclical factors directly has obvious and steady multi-scale features. The other quantity such as wind has less obvious and steady multi-scale characters. Weather changes in the scale analysis showed mutations in small-scale signal, but also in large-scale features on the local enhanced or suppressed information. Both high-resolution groups and low-resolution groups have redundancy resolutions in numerical model. Redundancy resolution depends on the scale feature of simulation objects. It is redundancy resolution except resolutions which nearly 50km,20km and 3km in this research and the simulation results are not changed significantly in redundancy resolution. Moreover, credibility of simulation results are lower and lower with the raise of resolution, simply raise the level of resolution to improve the model simulation ability is still limited. The model's ability to simulate free atmosphere are stronger than lower atmosphere on every experiment schemes. The numerical models also have a better ability to simulate physical quantity such as temperature which has a stronger multi-scale features compared to the quantity such as wind which has weaker multi-scale characters on every resolution and parameterization experiment schemes. There have close simulation results between schemes which have similar theory foundation. Experiment schemes show various degree of real atmospheric multi-scale movement's specific character in higher resolution experiment. However, the whole program simulation ability is generally not strong, the simulation capabilities of the boundary layer wind still weak, but the day of the simulation results is relatively good at night. The simulated wind profile evolution with time is significantly more intense than the measured and simulated wind speeds are generally larger than the real. Boundary layer simulation results in larger scale characteristics of the signal were stronger than the observed characteristics of the signal in the same scale, while smaller-scale signal is weaker than the actual observation.
|
PDF Full Text Request