YSU与MYNN边界层方案的黄海海雾模拟效果比较

发布时间：2018-01-21 14:31

本文关键词： 海雾黄海 WRF模式 YSU边界层方案 MYNN边界层方案　出处：《中国海洋大学》2014年硕士论文　论文类型：学位论文

【摘要】：海雾是发生在海上大气边界层内的一种天气现象，边界层方案所描述的湍流过程是影响海雾数值模拟的关键因素之一。本文旨在基于WRF（Weather Researchand Forecasting）模式，研究其YSU与MYNN边界层方案对黄海海雾的模拟效果，以便加深对它们的认识并试图对其不足加以改进。本文选取2次典型的黄海海雾个例为研究对象，它们分别是发生在2006年3月的一次浓厚海雾与发生在2012年3月一次较为淡薄的海雾。首先利用MTSAT卫星反演数据、再分析数据与常规观测数据，对海雾演变特征、海雾天气形势与边界层特征进行了详细的观测分析。然后设计并实施了一系列WRF数值试验，细致地对比分析了YSU与MYNN方案对黄海海雾的模拟效果。结果表明：1）对于2006年海雾个例，海雾边界层内湍流较强，湍流动量交换系数（Km）约为10m2/s，YSU模拟效果明显优于MYNN；因为MYNN边界层方案模拟的近海面湍流过强，Km达到了40m2/s，导致海雾逐渐脱离地面，不利于海雾的发展与维持，而YSU方案刻画的湍流强度适合海雾发展与维持。2）对于2012年海雾个例，海雾边界层内湍流较弱，雾区Km约为5m2/s，YSU的模拟效果明显不如MYNN；MYNN边界层方案刻画的湍流强度适合海雾发展与维持；而YSU方案则不能体现弱湍流，其给出的Km约1m2/s，导致模拟失败。3）YSU方案适合浓厚海雾的模拟，而MYNN方案适合淡薄海雾的模拟；适合海雾发生发展的湍流交换系数的平均值约为5m2/s左右。针对MYNN方案在强湍流的情况下模拟的近海面湍流过强导致海雾消散，而YSU方案在弱湍流的情况下对海雾的模拟效果不佳的问题，本文对MYNN与YSU方案进行了一系列的敏感性试验。我们试图改进MYNN方案对Case-2006与YSU方案对Case-2012的模拟效果，得到如下结果：1）对于Case-2006，MYNN湍流交换系数过大的情况下设计的敏感试验表明，直接将MYNN方案的湍流交换系数降低之后，对Case_2006模拟得到的非接海雾区虽然较原来有所减小，但是接海雾区有所改进；通过减小MYNN方案中与浮力项相关的系数降低湍流交换系数的试验，模拟得到的雾区明显增大，雾区改进显著；而对其他闭合参数修改的试验，模拟得到的雾区改变不大，不能提高海雾的模拟效果。2）对于Case_2012，YSU湍流交换系数过小的情况下设计的敏感性试验表明，直接将YSU方案交换系数增大后对Case_2012进行模拟，虽然雾区内的交换系数增大且接近模拟较好的MYNN方案的数值，但是雾区外的交换系数比MYNN增大一倍，导致海雾消散，模拟效果并没有得到改善；修改YSU方案中湍流交换系数的p指数虽然能够使得湍流交换系数增大，但对雾区模拟并没有改进，模拟得到的雾区几乎没有变化。
[Abstract]:Sea fog is a weather phenomenon occurring in the boundary layer of the sea atmosphere. The turbulent process described by the boundary layer scheme is one of the key factors affecting the numerical simulation of sea fog. Weather Researchand forecasting) mode. This paper studies the simulation effect of Huang Hai sea fog by its YSU and MYNN boundary layer schemes in order to deepen their understanding and try to improve their shortcomings. In this paper, two typical examples of Huang Hai sea fog are selected as the research object. They were a dense sea fog in March 2006 and a lighter one in March 1st 2012. First, the data were retrieved by MTSAT satellite. The characteristics of sea fog evolution, sea fog weather situation and boundary layer are analyzed in detail. Then a series of WRF numerical experiments are designed and implemented. The simulation results of Huang Hai sea fog by YSU and MYNN schemes are compared and analyzed in detail. The results show that the turbulence in the sea fog boundary layer is stronger for the case of 2006 sea fog. The turbulent momentum exchange coefficient (Km) is about 10 m ~ (2 / s) and the simulation effect of YSU is obviously better than that of MYNN. Because the MYNN boundary layer scheme simulates the strong turbulence over 40 m2 / s, the sea fog is gradually detached from the ground, which is not conducive to the development and maintenance of sea fog. The turbulence intensity described by the YSU scheme is suitable for the development and maintenance of sea fog. For the case of sea fog in 2012, the turbulence in the sea fog boundary layer is relatively weak, and km in the fog region is about 5 m2 / s. The simulation effect of YSU is obviously inferior to that of MYNN. The turbulent intensity described by the MYNN boundary layer scheme is suitable for the development and maintenance of sea fog. But the YSU scheme can not reflect the weak turbulence. The km of the scheme is about 1m2 / s, which leads to the failure of simulation. 3YSU scheme is suitable for the simulation of thick sea fog, while the MYNN scheme is suitable for the simulation of light sea fog. The average turbulent exchange coefficient suitable for sea fog development is about 5m2 / s. In view of the problem that the near sea turbulence simulated by the MYNN scheme is too strong in the case of strong turbulence, the sea fog dissipates, but the simulation effect of the YSU scheme on the sea fog is not good in the case of weak turbulence. In this paper, a series of sensitivity tests on MYNN and YSU schemes are carried out. We try to improve the simulation effect of Case-2006 and YSU schemes on Case-2012. The following results are obtained as follows: 1) the sensitive experiments designed for Case-2006 MYNN show that the turbulent exchange coefficient of the MYNN scheme is reduced directly. Although the non-sea fog area obtained by Case_2006 simulation is smaller than the original one, the sea fog zone is improved. By reducing the coefficients related to buoyancy in the MYNN scheme and reducing the turbulent exchange coefficient, the simulated fog region is obviously increased, and the fog zone is improved significantly. But for the other closed parameter modification test, the simulated fog area is not changed much, which can not improve the simulation effect of sea fog. 2) for Case_2012. The sensitivity experiments designed under the condition that the YSU turbulence exchange coefficient is too small show that the Case_2012 is simulated by increasing the exchange coefficient of the YSU scheme directly. Although the exchange coefficient in the fog region increases and is close to that of the simulated MYNN scheme, the exchange coefficient outside the fog zone is twice that of MYNN, resulting in the dissipation of sea fog, and the simulation effect is not improved. The modified p exponent of the turbulent exchange coefficient in the YSU scheme can increase the turbulent exchange coefficient, but there is no improvement on the simulation of the fog zone, and there is almost no change in the fog region obtained by the simulation.
【学位授予单位】：中国海洋大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：P732

【参考文献】