Experiment Of The Squall Line Ensemble Forecast Based On Improved Breeding Of Growing Mode

Mesoscale system’s small spatio-temporal scale, large meteorological elements gradient usually cause disasters such as thunderstorm, strong wind and hail. So it seriously affects people’s life and property, and the accurate forecast is of great significance, now we mainly adopt the ensemble forecast method to forecast. Due to the mesoscale ensemble forecast starts late in our country, there still exists some problems. Including the understanding of the distribution of initial error is not mature, so the perturbation solution needs improvement. Most scholars study one kind of solution, without fully considering the respective advantages and disadvantages of each scheme to be a reasonable match, etc. In this paper, in view of the squall line system, by using WRF model we firstly analyzing the error characteristics. Then through the test to determine the matching of different perturbation solutions in diverse resolution grids. Again according to the error’s characteristics, BGM(breeding of growing mode) method is put forward to improve constantly. Finally we have designed a suitable initial perturbation scheme. The main conclusions are as follows:(1) By analyzing the evolution of the error varying with time, we find the error of potential temperature and wind concentrats in wet convection and large values of convective available potential energy(CAPE), and the area becomes widened with the increasing of time. Thus, the area where error rapidly grows lies in wet convection and large values of CAPE.(2) Fully considering each perturbation solution’s characteristics, we know various BGM schemes in low resolution grid with root mean square error dynamic adjusting scale lagging average method(SLAF) in high resolution grid is better than being coupled with the mesoscale Barnes filter method judging from the dispersion and root mean square error. The U、 V、 T、 Qvapor’s field has obvious differences between each member, so it contains bigger possibility of the true evolution of atmosphere state. Meanwhile the ensemble average’s error is small.(3) According to the characteristic which is the error closely related with the moist convection area and CAPE distribution, in view of the disadvantage of the real error probability unknown during the dynamic breeding process, we bring forward three kinds of improved solutions about BGM scheme based on the moist convection area and comprehensively considering CAPE. Based on various evaluation indices of ensemble forecast system, the scheme which adjusted with6-hour precipitation comprehensively considering CAPE distribution behaves better. It improves the disadvantage of control forecast that heavy rain forecast range is too large, so it gets higher precipitation score than the others in heavy rain which shows that this scheme has certain improvement of the strong precipitation.From a concrete physical quantity field forecast such as radar reflectivity and wind profile, ensemble forecast denotes a clearer rainfall area, it also improvs the accuracy of westerly wind and north wind forecast that’s why the effect of the squall line forecast improved.