Validation of AOGCMs capabilities for simulation length of dry spells under the climate change and uncertainty in Iran

Precipitation is one of the vital climatic parameter, that plays a major role in human life.thereforethe impact of Precipitation in occurrence or non-occurrence of droughts and dry spells have been very effective.Identification and extraction length of dry spells in arid and semi-arid regions is very important. According to the most recent climate classification that has done, about 90 percent area of Iran is located in arid and semi-arid climate and more than 40 percent area of Iran is facing a severe water crisis.Therefore, understanding the behavioral mechanisms of dry spells have a great significance in arid and semi-arid Area, like country of Iran. Especially with the pose of the phenomenon of climate change that caused the worsening dryness and desertification in some of the region.Many research simulated dry spells with climate change approach and use of the output of AOGCM models. Researches in this category is less number, but the most recent research has been done by the authors(hashmy titles et al., 2015), investegation and modeling length dry spells in the southwest area of Iran.The aim of this research is examined,Validation of AOGCMs Capabilities for Simulation Length of Dry Spells under the Climate Change and Uncertainty in Iran
According to aim this research we used two database in this study: The first database involves collecting and analyzing information data base all (minimum temperature, maximum temperature, rainfall and sunshine) per dayly scale in 234 synoptic stations (with different statistical period), but the format for the data station and point during the period of statistical modeling was needed for more than 30 years, which has a large statistical defects were excluded, and finally 45 synoptic stations that have favorable conditions (the maximum area coverage and data continuous and reliable) who were selected for the final processing of the data first base. According that during the period is not uniform; therefore the period 1981-2010 was used as the base period. The second database contains data provided by version 5 models (LARS-WG) and on emissions scenarios (B1, A1B, A2) from AOGCM models for the decade 2050 to be downscaled. In fact, this data is the first data base (minimum temperature, maximum, precipitation and sunshine) prepare based on the format models for analysis and predict climate change, after downscale it.
Because this study was based on study and extraction length of dry spells in the range of long-term with approach of climate change, so the methodology based on several stages: In the first verification (validation) of LARS-WG, to ensure efficiency in the process model simulation will be discussed. Then assessment performance and capabilities 15 AOGCM models that in the new version Lars-wg. At the end of the precipitation threshold is defined and extract longest of length dry spells and compare it with the maximum length of the dry spells will be simulated.
After calibrating the model of statistical properties (comparison tests T, F and Pvalue (decision criteria), all stations were used to confirmed the validity of the model. The results of this calibration indicates that in more than 96% of stations for the minimum and maximum temperature, and sunshine model show high accuracy (in Dezful stations and Gorgan results of error was greater). In all of these stations were like Abadan station variables significant (P-value) at /05. It is acceptable that the data generated is random. Considering the bias error more than 95 percent of stations were very good agreement between the observed and modelling data (for every 4 variables).
Based on the the principles of (1 to 3), and using statistical methods and indicators, the AOGCM models to simulate and extraction during dry spells were examined and it was found that two models (Hadcm3 and GFDL-CM2.1) maximum performance (correlation) and the lowest error in estimating for simulation data precipitation. The model (INM-CM3 and NCPCM) have least amount of correlation and efficiency.
To estimate the maximum length of dry spells used Hadcm3 results under scenario (A2 and B1) for the decade 2050 and was skipped use the results of other models in this research.
Maximum dry spells in Iran, compliance with drynes condation in central and eastern areas. So that the country could be on the threshold of ./1 mm divided into 6 orbital region of the northern circuit during the period of 37 days (in the Rasht station) minimum and 351-day observation period southeastern Chabahar stations. The values show that the threshold of ./1 mm at more than 65 percent of the area’s dry spells over 7 months there was no rain on them yet. With a threshold of 5 mm needs maximum length of dry spells that lasted about a year with 364 days in Yazd station. That is roughly the size of 5 mm precipitation a year not registered at this station.
Modeling dry periods with calculate scenarios of climate change and the Considering Uncertainty resources of output models (AOGCM), showed that the worst-case scenario (A2), and critical status (2080), the average temperature of the country 2.7 degree (ºC) and Despite increased precipitation in some Stations, but 33% reduction in the all area of Iran. According to The most optimistic (B1) scenarios, increased the average of temperature 1.4 (ºC) and precipitation is Reduced 14% compared with the observation. The results of the uncertainty for dry Spells in 2050 and 2080 decades showed that in the two decades based on three scenarios (B1, A1B, A2), Length of Dry spells increased in all Area of Iran.Most of the changes in Length of Dry spells is, belongs the northwest Area of Iran (Oromeyeh, Khoy, Kermanshah, Hamedan and Lorestan).