Downscaling Of Temperature And Precipitation In The Daling River Basin,Liaoning Province

Projection of future climate changes is one of the hot issues in climate change research.The global climate models(GCMs)are important tools for future climate projection.However,the performances are poor at regional scale due to the coarse spatial resolution.Reanalysis data and statistical downscaling methods can effectively improve the reliability of GCMs.The global climate models have systemic errors,which increase the uncertainty of climate change projection.However,the multi-model ensemble average method can reduce the impact of model system errors to some extent and improve the credibility of future climate change scenarios.Based on the daily temperature and precipitation data(1979-2017)from nine meteorological stations around Daling River Basin,0.25° x 0.25° ERA-Interim grid data in 1979-2014,and eight CMIP6 models(1979-2014)and scenario data(2030-2099),this study applied Quantile Mapping(QM),Daily Translation(DT),Local Intensity Scaling(LOCI)and Delta,four statistical downscaling methods to downscale the historical temperature and precipitation of CMIP6 models in Daling River Basin.The best downscaling method and CMIP6 models are selected for temperature and precipitation using Taylor diagram.The four scenarios of SSP585,SSP370,SSP245,and SSP126,are downscaled based on the multi-model ensemble forecasts with respect to temperature and precipitation in the Daling River Basin.The main conclusions are as follows:(1)For the whole basin,ERA-Interim reanalysis data can reflect the variation of temperature and precipitation in the Daling River Basin with a high reliability.The simulation of temperature is better than precipitation,and the simulation of mean temperature is better than maximum temperature and minimum temperature.At daily scale,the average correlation coefficient between ERA-Interim and observations for mean temperature,maximum temperature,minimum temperature,and precipitation are 1.00,0.99,0.99,and 0.69,respectively.These two datasets showed high consistency in temperature.At monthly scale,the minimum temperature bias is negative,which means that the ERA-Interim overestimates the observation for minimum temperature.The mean temperature and maximum temperature bias are positive and negative,respectively.The root mean square error between ERA-Interim and observation for mean temperature is small(between 0.70? and 1.16?),which is lower than maximum temperature and minimum temperature.The root mean square error for precipitation between ERA-Interim and observation from June to August is significant.However,the ERA-Interim can accurately reflect the annual variation trend.At seasonal scale,the correlation coefficient between ERA-Interim and observation for mean temperature is higher than 0.93,which is greater than maximum and minimum temperature.ERA-Interim is better in winter and spring than summer and autumn for precipitation.At annual scale,the goodness of fit between ERA-Interim and observation between mean temperature,maximum temperature,and minimum temperature are 0.91,0.88 and 0.80,respectively.The goodness of fit between these two precipitation datasets is 0.74.ERA-Interim cand reflect the interannual fluctuation of precipitation quite well.(2)The best downscaling methods for precipitation,mean temperature,maximum temperature,and minimum temperature in the Daling River Basin are QM,DT,Delta,and Delta,respectively.The minimum temperature downscaling performance is the worst.Taylor diagram showed that the simulation of precipitation was not as good as temperature.The GCM INM-CM5-0 and MIROC6 model were the worst for precipitation projection.Therefore,BCC-CSM2-MR?EC-Earth3?EC-Earth3-Veg?FGOALS-g3?MPI-ESM1-2-HR?MRI-ESM2-0 models were selected for precipitation projection.All eight models are selected for mean temperature,maximum temperature and minimum temperature.Based on the best statistical downscaling method and the multi-model ensemble simulations,the results showed that the CMIP6 multi-model ensemble averaged precipitation,mean temperature,maximum temperature and minimum temperature can reflect the observed temporal and spatial variations very well in 1979-2014.(3)The annual precipitation in the Daling River Basin in the future(2030-2099)is between 400 mm and 1000 mm with a dramatic fluctuation.The annual precipitation in the most years is higher than the historical records.In 2030-2099,the annual precipitation will continue to decrease from southeast to northwest of the basin.The annual precipitation and precipitation anomaly percentages are in the order:SSP585>SSP370>SSP245>SSP126.Precipitation in the western basin will increase significantly.From SSP126 to SSP585,the mean temperature warming rate is 0.07?/10 a,0.27?/10 a,0.46?/10 a,and 0.65?/10 a,respectively.For the maximum temperature,the warming rate is 0.13?/10 a,0.27?/10 a,0.43?/10 a,and 0.54?/10 a,respectively.For the minimum temperature,the warming rate is 0.12?/10 a,0.29?/10 a,0.48?/10 a,and0.59?/10 a,respectively.However,from SSP126 to SSP585,the radiative forcing only increased twice times,the warming rate of mean temperature,maximum temperature,and minimum temperature increased eight times,three times,and four times,respectively.In spatial,the future temperature will still gradually decrease from south to north of the basin.The anomaly ranges of mean temperature,maximum temperature and minimum temperature are 0?-2.2?,0.2?-2.4? and-1.0?-1.2?,respectively.The mean temperature and maximum temperature generally showed a warming trend in four scenarios,while the minimum temperature in SSP126 and SSP245 scenarios showed cooling trends,which is less than 1? compared with the historical period.