Search

The parameterization of physical processes is one of the critical elements to properly simulate the regional climate over eastern China. It is essential to conduct detailed analyses on the effect of physical parameterization schemes on regional climate simulation, to provide more reliable regional climate change information. In this paper, we evaluate the 25-year (1983-2007) summer monsoon climate characteristics of precipitation and surface air temperature by using the regional spectral model (RSM) with different physical schemes. The ensemble results using the reliability ensemble averaging (REA) method are also assessed. The result shows that the RSM model has the capacity to reproduce the spatial patterns, the variations, and the temporal tendency of surface air temperature and precipitation over eastern China. And it tends to predict better climatology characteristics over the Yangtze River basin and the South China. The impact of different physical schemes on RSM simulations is also investigated. Generally, the CLD3 cloud water prediction scheme tends to produce larger precipitation because of its overestimation of the low-level moisture. The systematic biases derived from the KF2 cumulus scheme are larger than those from the RAS scheme. The scale-selective bias correction (SSBC) method improves the simulation of the temporal and spatial characteristics of surface air temperature and precipitation and advances the circulation simulation capacity. The REA ensemble results show significant improvement in simulating temperature and precipitation distribution, which have much higher correlation coefficient and lower root mean square error. The REA result of selected experiments is better than that of nonselected experiments, indicating the necessity of choosing better ensemble samples for ensemble. [ABSTRACT FROM AUTHOR]

Strong logical relationships exist between temperature difference and hydro-meteorological variables such as precipitation, dry days, and dry spells. These relationships can be used for knowing how the differences of total and top 10% heavy precipitations, dry days, and dry spells will change with respect to temperature differences? Here, daily precipitation and temperature records over Turkey are considered from 1971 to 2010, inclusive. First 30-year (1971-2000) data are used for training for temperature difference function (ΔT) and the remaining 10-year (2001-2010) is reserved for validation. The temperature difference function is validated by mean relative error, and the results are calculated for total and top 10% heavy precipitation, dry days, and spells as 14.60%, 6.14%, 2.81%, and 11.89%, respectively. The strong relationship exists between temperature difference function and standard deviation of corresponding climatic variables. Results show that standard deviation and mean relative error (MRE) have linear correlation. Also, ΔT model predictions are compared with CMIP5 RCPs projections from 2020 to 2100. The suggested model predictions fall within the range of various atmosphere-ocean global circulation (climate) model (AOGCM) scenario RCP 4.5 and RCP 8.5 projections, in addition to model predictions cross-correlation for RCP 2.6 and RCP 6.0. [ABSTRACT FROM AUTHOR]

This study investigates the capabilities of two regional models (the ICTP RegCM3 and the climate version of the CPTEC Eta model – EtaClim) in simulating the mean climatological features of the summer quasi-stationary circulations over South America. Comparing the results with the NCEP/DOE reanalysis II data it is seen that the RegCM3 simulates a weaker and southward shifted Bolivian high (BH). But, the Nordeste low (NL) is located close to its climatological position. In the EtaClim the position of the BH is reproduced well, but the NL is shifted towards the interior of the continent. To the east of Andes, the RegCM3 simulates a weaker low level jet and a weaker basic flow from the tropical Atlantic to Amazonia while they are stronger in the EtaClim. In general, the RegCM3 and EtaClim show, respectively a negative and positive bias in the surface temperature in almost all regions of South America. For both models, the correlation coefficients between the simulated precipitation and the GPCP data are high over most of South America. Although the RegCM3 and EtaClim overestimate the precipitation in the Andes region they show a negative bias in general over the entire South America. The simulations of upper and lower level circulations and precipitation fields in EtaClim were better than that of the RegCM3. In central Amazonia both models were unable to simulate the precipitation correctly. The results showed that although the RegCM3 and EtaClim are capable of simulating the main climatological features of the summer climate over South America, there are areas which need improvement. This indicates that the models must be more adequately tuned in order to give reliable predictions in the different regions of South America. [ABSTRACT FROM AUTHOR]