Regional Temperature Modeling of Iran using PRECIS Regional Climate Model, Case Study: 1976-1990

IntroductionSince the beginning of industrial revolution and especially in the recent half of 20Th century, it is increasingly recognized that the magnitude of human influence on the Earth persistently intensified. As the first time in the history of our planet, emissions of trace gasses from human activities has been equaled, or even exceed from natural sources (Henderson et al 2001). Human influences will continue to change the atmospheric composition throughout the 21st century. (Jones et al 2003). Anxieties are now plausibly felt, and extensively uttered about the possible immense of impact on the global climate owing to the enhanced level of greenhouse gases concentrations on the atmosphere (Fischer et al 2002). The intergovernmental Panel on Climate Change (IPCC) reported that the global mean temperature has been increased 0.6oC during 20th century while the atmospheric concentration of carbon dioxide also increased from 280ppm to 370 ppm in third Assessment Report (TAR) published in 2001. Regional climate models of RegCM3 and PRECIS are performed for Carpathian basin assessment during the periods of 1961-1990 & 2071-2100. It is found that seasonal mean temperature simulation field, expect winter, are slightly underestimated by RegCM in Hungary. But they are overestimated by PRECIS simulation. Except autumn, other seasonal precipitation fields over Hungary are usually overestimated by RegCM simulation with the largest bias values in spring (Bartholy et al 2009). Climate change in the past decade in Jianghuni valley is studied by using statistical downscaling techniques. Both frequency and strength of extreme climate events such as hot weather, droughts and floods have increased remarkably since 1990s. PRECIS also is used to provide future climate prediction over the valley. The results give an average surface warming of 2.9oC under the SRES B2 emission scenario by the end of this century (2071-2100). They found that precipitation may increase on the same period (Tian et al 2006). Regarding to the role of climate change simulation in agricultural, water resources and other economical sectors, the skill of PRECIS regional climate model in simulation of the mean monthly temperature of Iran during 1961-1990 has been studied in this research.Methodology and DataIn this paper PRECIS regional climate model that has been developed by Hadley Center of United Kingdom Meteorology Office are used. PRECIS is a hydrostatic version of the full primitive equation, i.e. vertical acceleration in the atmosphere is assumed to be smaller than hydrostatic equilibrium and hence vertical motions are diagnosed separately from equations of state. It has a complete representation of the Carioles force and employs a regular latitude- longitude grid in the horizontal and a hybrid vertical coordinate. A terrain following σ-coordinate is considered at the lower four levels with purely pressure coordinate at the top three levels. The model has 19 vertical levels in the atmosphere and 4 levels in the soil. The PRECIS RCM model can be run at two different horizontal resolutions of 0.44°×0.44° and 0.22°×0.22°. The validation of model has been done by comparing observation data and model output data with two different methods of region to region and station to station. In this study, the model domain covers 23 to 45 degree in north and 43 to 68 degreeineast and its horizontal resolution is 50 x 50 km. The HadAM3P global data set is used to drive the PRECIS model. The horizontal resolution of the HadAM3P boundary data is 150 km. It covers the period of 1960-1990 and 2070-2100 respectively (Wilson et al 2005). The first year in each PRECIS experiment is considered as a spin-up period and these data are not used in any analysis. Period of study in this work is 15 years including 1976-1990. Discussion and Results Masoudian temperature zoning of Iran (Masoudian 2008) is considered for computing regional bias of the model simulations. In this approach, temperature regions of Iran have been categorized in 6 regimes of coldcool, mild, semi- warm, warm and hot. So, monthly to annual bias and standard deviation of the model outputs have been calculated using region to region method. Table 1 show that minimum and maximum annual biases have been found in cold and mild regimes with -0.3 and -2.2°C, respectively. Statistical analysis confirmed that mean and standard deviation of the observed and modaeled data are same. In this regard, F and t statistical tests show that there is no significant difference between modeled and observed data.ConclusionAs a main result, PRECIS skill in modeling regional temperature and mean temperature of Iran is well. Minimum and maximum biases of modeling over the country are -3.2 for January and 0.1°C for April and September. At monthly modeling, Validation of standard deviation shows that modeling without Sulfur cycle has lower error than modeling with Sulfur cycle included. There is no significant difference between PRECISmodeled data and actual data retrieved from weather stations. So, as a powerful regional model, PRECIS can be used for regional climate modeling over Iran and future climate change projections.