Simulation And Projection Of Summer Daily Precipitation Based On Nonhomogeneous Hidden Markov Model Over The Yangtze-Huaihe River Basin

The nonhomogeneous hidden Markov model (NHMM) of statistical downscaling method is introduced in this paper, based on the observed summer daily precipitation data at 56 meteorological stations in the Yangtze-Huaihe River Basin, and the ERA-40 reanalysis data of European Centre for Medium-Range Weather Forecasting(ECMWF) during 1961-2002, the NHMM downscaling model is established and the simulation capabilities of NHMM for summer daily precipitation over the east monsoon area in China(the Yangtze-Huaihe River Basin as the representative) have been assessed, then the simulation results have been compared between NHMM downscaled and BCC-CSM1.1(m), IPSL-CM5A-MR, MPI-ESM-MR.Finally, the atmospheric circulation data form 3 climate models which mentioned above are used to drive the NHMM, simulating the change of summer precipitation in the early(2016-2035), middle(2046-2065) and late period(2081-2100 year) in the 21st century over the Yangtze-Huaihe River Basin under the representative concentration pathway 4.5 scenario. The results show that:(1) Based on the interannual correlation coefficients between atmospheric circulation variable and the summer precipitation over the Yangtze-Huaihe River Basin, the simulation results from different NHMM with different circulation variable have been compared. Found that the combination of 500hPa geopotential height, sea level pressure,500hPa relative humidity and 500hPa zonal wind is the best predictor of NHMM over the Yangtze-Huaihe River Basin.(2) The evaluation of the simulation ability of NHMM in the independent test indicates that:NHMM performs well in simulating daily precipitation over the Yangtze-Huaihe River Basin by establishing the relationship between the transition probabilities of the precipitation probability distribution states and the synoptic-scale atmospheric predictors. The simulated probability distribution function(PDF) curves are close to the observations with the Brier score (BS) less than 0.11% and the Significance score(5score) greater than 0.84; The relative errors of summer total precipitation, number of rainy days(≥1mm), number of rainy days for daily precipitation more than 10mm, simple daily intensity,95th percentile value of precipitation are less than 10% and the spatial correlation coefficients are greater than 0.75; The correlation coefficients of area-averaged precipitation indices’annual sequences between simulations and observations are greater than 0.62.(3) The NHMM can effectively improve the climate model’s capabilities in simulating the summer daily precipitation over the Yangtze-Huaihe River Basin. After downscaling, the median values of BS are reduced by 0.55%,0.29%,0.03%, 0.15% and the median values of Sscore are increased by 0.16 and 0.23,0.22,0.07 relative to the simulation results from BCC-CSM1.1(m), IPSL-CM5A-MR, MPI-ESM-MR and multi-model ensemble, which means that the simulated probability distribution curves are closer to the observations. The relative errors of precipitation indices are generally below 10% and the spatial correlation coefficient increased from less than 0.6 to 0.8 above. The root mean square errors are less than 0.75, which means the downscaled results are closer to the observations.(4) Under the RCP4.5 scenario, the summer precipitation over the Yangtze-Huaihe River Basin will increase generally in the future. According to the simulation results from multi-model downscaling ensemble:the area-averaged summer total precipitation will increase 9.57%,17.43%,19.16%; the number of rainy days will increase 7.65%,13.54%,14.92%; the simple daily intensity will increase 1.77%, 3.37%,3.52%; the number of rainy days for daily precipitation more than 10mm will increase 9.21%,16.61%,18.54%, the 95th percentile value of precipitation will increase 2.11%,3.72%,3.85%, respectively relative to the current climate scenarios(1986-2005) in the Early, middle and late in the 21st century; And the consecutive dry days will decrease-8.35%-10.44% and-11.10%, respectively.(5) The spatial distribution of the relative change (relative to 1986-2005) of precipitation indices under the RCP4.5 scenario shows that:the summer total precipitation, number of rainy days, simple daily intensity, number of rainy days for daily precipitation more than 10mm,95th percentile value of precipitation will increase at most stations, which means that the extreme precipitation will increase and the consecutive dry days will decrease relative to the current climate scenarios. The relative change of each index will gradually increase from early to late in the 21st century. And the changes of summer total precipitation, number of rainy days, number of rainy days for daily precipitation more than 10mm are greater than consecutive dry days, simple daily intensity and 95th percentile value of precipitation, which means that the precipitation frequency will be bigger. In addition, the difference between stations of summer precipitation over the Yangtze-Huaihe River Basin will increase in the future, shows that the regional features of precipitation will be stronger in the future.