A Method of Creating Initial Perturbation Based on LAF Method and Experiments of Ensemble Forecast
Lifeng Zhang, Niu Zheng-yu
Abstract
Lifeng Zhang, Niu Zheng-yu
Abstract
An improved LAF method is brought forward based on LAF idea.The idea of this new method is adding different initial perturbations to model initial field for different perturbation forecastings,which generates members of ensemble forecasting.The perturbation field is derived from difference between different analysis fields on lagged various time.Thus,not only the perturbation reflects the difference between different analysis fields but also lagged various time embodies idea of LAF method.Then,ensemble forecasting experiments of the first Meiyu rainstorm in Anhui province from 14 to 15 June 2004 are made respectively with LAF method and advanced LAF method with AREM model used.The results show that advanced LAF method is superior to classical LAF about forecasting rainband position and intensity of the rainfall.After further analyzing,it is found that to the same perturbation the forecasting effect of adding perturbation forecast is better than that of subtracting perturbation forecast.Even the ensemble forecast gets much closer to the observation while only the results of adding perturbation forecast are averaged.
OpenAlex reports 1 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
An improved LAF method is brought forward based on LAF idea.The idea of this new method is adding different initial perturbations to model initial field for different perturbation forecastings,which generates members of ensemble forecasting.The perturbation field is derived from difference between different analysis fields on lagged various time.Thus,not only the perturbation reflects the difference between different analysis fields but also lagged various time embodies idea of LAF method.Then,ensemble forecasting experiments of the first Meiyu rainstorm in Anhui province from 14 to 15 June 2004 are made respectively with LAF method and advanced LAF method with AREM model used.The results show that advanced LAF method is superior to classical LAF about forecasting rainband position and intensity of the rainfall.After further analyzing,it is found that to the same perturbation the forecasting effect of adding perturbation forecast is better than that of subtracting perturbation forecast.Even the ensemble forecast gets much closer to the observation while only the results of adding perturbation forecast are averaged.
Key concepts: Perturbation (astronomy), Mathematics, Computer science, Meteorology, Applied mathematics, Statistical physics, Physics, Quantum mechanics