A Random Physical Perturbation Method Based on Energy Conservation

Abstract

The invention relates to an energy conservation based random physical perturbation method, comprising the following steps of: 1) according to actual needs, selecting a numerical forecasting mode and determining the position of a final tendency item of each parameter in a mode forecasting equation; 2) after determining the position of the parameter, according to a perturbation form, determining a perturbation scheme of the tendency item by integrating, wherein an expression of the perturbation form is defined in the specification; 3) performing two-dimensional Fourier decomposition on epsilon (t) in a horizontal direction, filtering away high-frequency noise, assigning 0 to a coefficient of a wave with the wavelength 1-4 times the grid distance, and finally performing inverse Fourier transform on the regulated coefficient to synthesize new epsilon (t); and 4) repeating the step 3) until a perturbation coefficient corresponding to each variable is obtained, and substituting the perturbation coefficient to the expression of the perturbation form, wherein the perturbation coefficient is defined in the specification. The method has the beneficial effects that: a completely independent physical process is subjected to tendency perturbation for ensuring energy conservativeness of the tendency perturbation of the mode variables, so that the method can improve the ensemble dispersity of the mode variables in a boundary layer and the forecasting level of storm-scale ensemble forecasting can be remarkably improved.

Description

technical field [0001] The invention relates to the field of weather forecast, in particular to a random physical disturbance method based on energy conservation. Background technique [0002] A series of studies in the past have shown that storm-scale ensemble forecasting is feasible and effective, but the perturbation methods of various storm-scale ensemble forecasting systems still need to be perfected. For example, the storm-scale ensemble forecast system of the Center for Storm Analysis and Prediction (CAPS) in the United States can improve the forecast score by increasing the number of members and resolution, but its ensemble disturbance scheme is fixed, that is, the initial disturbance of each ensemble member adopts a fixed disturbance Amplitude, and matched with a specific physical scheme; the lateral boundary conditions of some ensemble members are provided by the short-term forecast of the NAM model, and the same lateral boundary conditions among members will limit...

AI Technical Summary

Problems solved by technology

[0003] Because the evolution of the initial error over time is significantly different in baroclinic instability and convective instability, the initial disturbance constructed by the medium-term ensemble disturbance method cannot grow rapidly in the convective system, that is, the initial disturbance structure is not consistent with the development of the storm system. On the other hand, the incompatibility between the initial disturbance of the storm scale and the scale of the side boundary disturbance will also limit the growth of the initial disturbance and the dispersion of the set

Multi-physics processes and multi-model ensembles help to improve the dispersion problem, but the precipitation forecast variables in the micro-physics scheme of the model and the threshold condition judgment of the micro-physics process are different, and have a significant impact on the precipitation intensity and fall area forecast. For different types of storm systems that occur in real time, it is blind to use a fixed set configuration to select different modes of microphysics schemes

Images