Application On Climate Dynamical Downscaling Skills Using Nudging Method Over China

The nudging technique is an important method for improving dynamical downscaling and is sensitive to nudging parameters.However,most studies still adopt the default nudging schemes.It is significant to study the sensitivity of the nudging parameter to achieve better downscaling result.The Qinghai-Tibet Plateau(TP)has attracted much attention due to its role in global and regional atmospheric circulation and its sensitivity to climate change.Research on TP needs adequate meteorological data.However,existing observation data are scarce and unevenly distributed over time and space.The resolution of the reanalysis data is coarse and its applicability is limited.The establishment of a high-resolution data set for TP using dynamical downscaling is an effective way to improve this situation.Based on this,we carried out a series of studies on the sensitivity of nudging parameters to obtain the suitable nudging scheme,then used it for the dynamical downscaling process to produce two sets of climate data with different resolutions,and used the data to develop the TP analysis.In this paper,the Grid Nudging and Spectral Nudging were applied to the WRF(Weather Research and Forecasting)model,and the FNL data was used as the driving field to simulate the 2009 and 2010 summer climate over Chinese Mainland.The performances of simulations with varying nudged variables,levels,nudging time and cut-off wavelength were compared.The results showed all nudged variables influenced the simulation of meteorological elements near the surface and at 500 hPa.Of Grid Nudging,nudging potential temperature(?)had the largest improvement to the simulation of meteorological elements,followed by nudging horizontal wind(uv),and nudging water vapor mixing ratio(q).Considering the simulations over the whole country,in view of nudging time,1 h or 3 h was the optimal nudging time for nudging uv,while it was 3 h for nudging ?,and 1 h for nudging q.In view of nudged level,the performance was best when the nudged level was above the PBL or 850 hPa for nudging uv or ? and above the PBL for nudging q.Of Spectral Nudging,nudging ? had the largest improvement to the simulation of meteorological elements,followed by nudging uv,and nudging geopotential height(ph).Considering the whole country,the performance was best when the cutoff wavelength was 500 or 1000 km for nudging uv and 1000 km for nudging ? or ph.Comparing with single nudged variable,the nudging of all variables had obvious advantages and performed better in almost all regions and levels.500 or 1000 km was also the optimal cutoff wavelength,when all variables were nudged.Using a one-way nested configuration,different nudging schemes were also compared by evaluating regional climate modeling over TP in January,April,July and October of 2009.Ultimately,the suitable nudging scheme for high-resolution climate simulation over TP was obtained.The results showed that for meteorological elements near the surface and at 500 hPa,the simulations were poor when nudging was absent,and improved with the presence of nudging.Improvement was greater when nudging scheme was used in both two nested domain.For regional climate modeling ove TP,the design of Spectral Nudging in outer domain and Grid Nudging in inner was recommended.Finally,a long simulation was performed using one-way nested WRF model to downscale FNL data over TP,resulting in two sets of high-resolution climate data.The spatial resolution of the data was 30 and 6 km,respectively.The synoptic circulation,precipitation,and 2 m temperature characteristics over TP were analyzed using data sets.The results showed that during the spring,autumn and winter the flat westerly flow dominated the TP,while in summer the flow from the Indian Ocean and the Bay of Bengal strongly affected the TP and Southwest China.The precipitation over the East TP was more than the one over the West.Precipitation was mainly concentrated in May-September.Within three years,the precipitation was largest in July 2009,and least in January 2010.Of high-resolution climate data and TRMM precipitation,their spatial distribution and temporal evolution are similar.The surface temperature was warm over the East TP and cold over the West.The maximum monthly mean temperature appeared in July,and the minimum in January.Within three years,the temperature was warmest in July 2010,and coldest in January 2011.Compared with the site temperature data,their temporal evolution is similar,and 2 m temperature after the lapse rate correction is very close to the site.