A Method for Refinement Processing of Numerical Prediction Grid Point Temperature Forecast Data

Abstract

The invention provides a refined processing method for temperature forecast data of a numerical forecasting grid point. The refined processing method specifically comprises the steps of obtaining experimental data after system initialization; reading longitude and latitude coordinates of an observation point and four grid points around, the live temperature data of the observation point and the temperature forecast values of the four grid points around; carrying out error analysis and data sieving; then calculating an error arithmetic mean value of each of the 24 solar terms and a sliding mean value of a same forecast time error of continuous three days; carrying out inverse tangent function optimization processing for the obtained values; then obtaining a weight coefficient of influences from the two errors abovementioned to the forecast result; then calculating sequentially based on a set sequence to obtain a set of an error sliding mean value of each observation point, a solar term error mean value and a weight coefficient; carrying out amended calculation for the temperature data of the numerical forecasting grid point in the county territory represented by each observation point by using the result abovementioned; and finally carrying out refined interpolation calculation, daily maximum temperature calculation and daily minimum temperature calculation for an appointed coordinate point by using the amended data.

Description

technical field [0001] The invention relates to the field of meteorological forecast, in particular to a method for finely processing numerical forecast grid point temperature forecast data. Background technique [0002] At present, the forecast methods used by meteorological departments for operational systems mainly include DMO method, PP method, artificial neural network method, MOS method, similarity forecast method, dynamic method and Kalman filter method, etc. [0003] The traditional DMO method is to analyze the forecast results of the model elements on the grid points to the specific stations through interpolation, and obtain the element forecasts on the stations. For the elements that are not directly output by the model, they are calculated by empirical formulas. The disadvantage is that the model error has no effect. Correction ability, the forecast accuracy is completely dependent on the model, and the accuracy of the element forecast is often not very high compa...

AI Technical Summary

Problems solved by technology

[0006] In the current situation where people require higher accuracy in temperature forecasting, the traditional methods can no longer meet the current demand for high-precision temperature forecasting.

Images