研究I的结果表明:线性平衡方程(LBE)在热带地区不适用,而进一步改进方向是削弱LBE在该区域的约束程度。本文以此为基础,在GRAPES(global/regional assimilation and prediction system)全球变分同化系统中引入动力与统计混合平衡约束方...研究I的结果表明:线性平衡方程(LBE)在热带地区不适用,而进一步改进方向是削弱LBE在该区域的约束程度。本文以此为基础,在GRAPES(global/regional assimilation and prediction system)全球变分同化系统中引入动力与统计混合平衡约束方案。新方案在逐层求解LBE的基础上增加垂直方向的线性回归,回归系数随纬度和高度变化。针对背景误差协方差的分析表明,新方案可以更好的保证独立分析变量间预报误差不相关的基本要求,并大幅度减小热带地区平衡气压预报误差方差的量值和占总方差的比例。单点试验结果表明,与LBE方案相比,新方案对中、高纬影响很小,但在热带地区成功实现了风、压场分析的解耦,两者分析更为独立。并且,虽未考虑具体波动模态,但新方案给出的风、压场协相关结构与研究I的理论分析结果相近。一个月的同化循环与预报结果表明,引入新方案后,赤道外地区的同化预报效果为中性偏正,而热带地区风场的同化预报效果显著提高,LBE方案中平流层低层的风场同化预报异常被基本消除。展开更多
本文基于GRAPES全球模式的短期预报误差样本，利用赤道波动正规模态研究了热带风、压场平衡特征，并根据这些特征分析了线性平衡方程（LBE）在该区域应用时存在的问题。结果表明：（1）赤道波动能成功解释热带短期预报误差样本的大部分...本文基于GRAPES全球模式的短期预报误差样本，利用赤道波动正规模态研究了热带风、压场平衡特征，并根据这些特征分析了线性平衡方程（LBE）在该区域应用时存在的问题。结果表明：（1）赤道波动能成功解释热带短期预报误差样本的大部分分量，对流层中层为60%～80%，对流层顶和平流层低层为80%以上。（2）在可解释的误差方差中，赤道罗斯贝波（ER）占比仅为30%～55%，其他赤道波动的作用不可忽视。（3）在ER模态基础上引入其他赤道波动会大幅削弱原有风、压场平衡约束，重力惯性波与Kelvin波的作用最为显著。此时，对流层中层位势高度h与u风、v风间的约束接近于零，而平流层低层h–u的平衡特征由Kelvin波主导。（4）LBE主要表达了 ER 模态下的风、压场平衡特征，与实际情形相比高估了热带风、压场的耦合程度，进一步的改进中需削弱这一虚假平衡，使得热带风、压场分析变得更加独立。展开更多
Buckling could be induced when shallow arches were subjected to vertical step loads. In-plane static and dynamic buckling of shallow pin-ended parabolic arches with a horizontal cable was investigated. Based on the eq...Buckling could be induced when shallow arches were subjected to vertical step loads. In-plane static and dynamic buckling of shallow pin-ended parabolic arches with a horizontal cable was investigated. Based on the equations of motion derived from Hamilton's principle, nonlinear equilibrium equations and static buckling equilibrium equations were deduced. Through the pseudo-static analysis, approximate solutions to the lower and upper dynamic buckling loads under step loads were obtained, for shallow parabolic arches. The results show that dynamic buckling and snap-through buckling are impossible when modified slenderness ratio λ<λc and λ>λs, where λc and λs denote critical slenderness ratios of bucking and snap-through buckling, respectively; effects of the stiffness of the horizontal cable on the dynamic buckling are significant; and the dynamic buckling loads under a equivalent central concentrated step load are lower than the loads under a distributed load appreciably.展开更多
文摘研究I的结果表明:线性平衡方程(LBE)在热带地区不适用,而进一步改进方向是削弱LBE在该区域的约束程度。本文以此为基础,在GRAPES(global/regional assimilation and prediction system)全球变分同化系统中引入动力与统计混合平衡约束方案。新方案在逐层求解LBE的基础上增加垂直方向的线性回归,回归系数随纬度和高度变化。针对背景误差协方差的分析表明,新方案可以更好的保证独立分析变量间预报误差不相关的基本要求,并大幅度减小热带地区平衡气压预报误差方差的量值和占总方差的比例。单点试验结果表明,与LBE方案相比,新方案对中、高纬影响很小,但在热带地区成功实现了风、压场分析的解耦,两者分析更为独立。并且,虽未考虑具体波动模态,但新方案给出的风、压场协相关结构与研究I的理论分析结果相近。一个月的同化循环与预报结果表明,引入新方案后,赤道外地区的同化预报效果为中性偏正,而热带地区风场的同化预报效果显著提高,LBE方案中平流层低层的风场同化预报异常被基本消除。
文摘本文基于GRAPES全球模式的短期预报误差样本，利用赤道波动正规模态研究了热带风、压场平衡特征，并根据这些特征分析了线性平衡方程（LBE）在该区域应用时存在的问题。结果表明：（1）赤道波动能成功解释热带短期预报误差样本的大部分分量，对流层中层为60%～80%，对流层顶和平流层低层为80%以上。（2）在可解释的误差方差中，赤道罗斯贝波（ER）占比仅为30%～55%，其他赤道波动的作用不可忽视。（3）在ER模态基础上引入其他赤道波动会大幅削弱原有风、压场平衡约束，重力惯性波与Kelvin波的作用最为显著。此时，对流层中层位势高度h与u风、v风间的约束接近于零，而平流层低层h–u的平衡特征由Kelvin波主导。（4）LBE主要表达了 ER 模态下的风、压场平衡特征，与实际情形相比高估了热带风、压场的耦合程度，进一步的改进中需削弱这一虚假平衡，使得热带风、压场分析变得更加独立。
基金Project (50478075) supported by the National Natural Science Foundation of China
文摘Buckling could be induced when shallow arches were subjected to vertical step loads. In-plane static and dynamic buckling of shallow pin-ended parabolic arches with a horizontal cable was investigated. Based on the equations of motion derived from Hamilton's principle, nonlinear equilibrium equations and static buckling equilibrium equations were deduced. Through the pseudo-static analysis, approximate solutions to the lower and upper dynamic buckling loads under step loads were obtained, for shallow parabolic arches. The results show that dynamic buckling and snap-through buckling are impossible when modified slenderness ratio λ<λc and λ>λs, where λc and λs denote critical slenderness ratios of bucking and snap-through buckling, respectively; effects of the stiffness of the horizontal cable on the dynamic buckling are significant; and the dynamic buckling loads under a equivalent central concentrated step load are lower than the loads under a distributed load appreciably.
