The Diagnosis And Prelimiary Improvement Of Cumulus Parameterization Schemes

In the numerical weather prediction models and climate models, the cumulusparameterization scheme is a crucial part of the impact model performance. So it ismeaningful to evaluate the performance of cumulus parameterization schemes andimprove them.First Using the GRAPES（Global/Regional Assimilation and PrEdiction System）Single Column Model and the GCSS WG4Case3data, numerical experiments wereconducted to evaluate the cumulus parameterization schemes in GRAPES, such asBMJ, SAS and KFeta. Then according to the results of reseach, a new modifiedversion of KFeta scheme is proposed. Finally, Because of the importance ofconvective momentum transport (CMT) and GRAPES do not include it, so weintroduced CMT into the GRAPES by including CMT part to the KFeta scheme. Andnumerical experiments are conducted to examine the sensitivity of the numericalsimulation of Hurricane Megi(2010) to the CMT. It is shown:(1) BMJ has stronger ability to transport water vapor upward which leads to the lowtroposphere drier, middle and high troposphere wetter, the atmosphere that adjustedby BMJ is too stable. KFeta is trend to over prediction of low-level moisture andproduced serious upper-level cool bias which was caused by excessive overshoot.(2) In the new modified scheme the vertical velocity equation is modified, namelyenhanced turbulent disturbance, include the effect of pressure disturbance, adjust theminimum of the maximum environment inflow rate and the precipitation efficiencycoefficient rate. Single-profile test, single-column tests and GRAPES_Meso practicalexamples show that the modified scheme can effectively improve the simulation ofhigh-level-layer potential temperature, the impact of precipitation and humidity fieldis smaller. And also the modified scheme proposed by Anderson et al.(2007) isexamined,numerical experiments shows that it can reduce high-level cold biaseffectively, but the impact is multifaceted and complex, more research is need.(3) Numerical experiments are conducted to examine the sensitivity of thenumerical simulation of Hurricane Megi(2010) to the CMT. Numerical experimentsshow that, CMT can effectively improve Megi’s path simulation. On the strength simulation, at time period Ⅰ (October14th-October18th), CMT inhibit thedevelopment of Megi, and in time period Ⅱ (October18th-October22th) theincrease in intensity of the typhoon. This may caused by the wind field is related withthe CMT. In the time period Ⅰ, the low-level wind field near the typhoon centercaused by CMT is mainly divergence or negative vorticity, and therefore inhibit thedevelopment of typhoon, and time period Ⅱ, the low-level wind field near thetyphoon center caused by CMT is convergence or positive vorticity, promote thedevelopment of the typhoon.