Development Of The Global Non-hydrostatic Model GRAPES_YY On The Yin-Yang Grid And The Performance Verification

Taking advantage of the quasi-uniformity of the Yin-Yang grid and reconstructing the dynamic core of the GRAPES global model on the Yin-Yang grid(GRAPES＿YY)can fundamentally solve a series of problems associated with the polar singularity,which is conducive to global high-resolution simulations.In this study,based on the aqua planet version,the GRAPES＿YY model is developed by coupling the Co LM land surface model and a subgridscale topographic gravity-wave drag parameterization scheme to complete the physical package for the application in medium-range numerical weather forecast.In order to improve the numerical accuracy and stability of the model,the upper boundary condition that satisfies the static balance,a vertical velocity damping algorithm and a non-interpolation algorithm of potential temperature in vertical direction are implemented into the dynamic core.The new model is then tested with a series of medium-range forecasting experiments,including a batch tests of global circulation and precipitation,a case analysis of a Meiyu front and an explosive cyclone development at high-latitude region with a horizontal resolution of 0.75°×0.75° for the verification of computational stability and dynamics-physics interaction.The results of the batch tests display reasonable distribution of global circulation and precipitation within the first three-day period.Approximately the first seven-day forecast is available with a correlation coefficient number above 0.6 for the geopotential height on 500 h Pa.Over estimation of the convective precipitation in low latitudes and large-scale precipitation in mid-latitude area,however,is revealed when the integration goes on.Besides the batch simulations,two case simulations is selected here to show the performance of the models.The rainfall simulated by the GRAPES＿YY performs more advantageous than the GRAPES-GFS model in a case simulation of Meiyu front.In the simulation of an explosive polar cyclone,the two models show similar prediction results of the path,intensity and precipitation in the first 66 hours.Unfortunately,the GRAPES＿YY displays slightly worse path of the cyclone movement than the GRAPES-GFS model later in the case simulation.Generally,the GRAPES＿YY model is capable of stable medium-range forecasting test,which basically does well in a resolution of 0.75°.