عنوان مقاله [English]
Investigating the components of general atmospheric circulation is important to discover the rules governing the country's rainfall. The relative vorticity as a key variable of synoptic motions is one of the best components in this regard. Relative vorticity is a measure of the intensity and direction of spin in a circular movement which is performed by a unit volume of air around the vertical axis perpendicular to the plane over which this rotation occurs. This research is looking for the relationship between spatiotemporal variations of relative vorticity and precipitation in Iran to improve forecasting skills.
In this study the connection between the 500 hPa relative vorticity field and precipitation over Iran is investigated using canonical correlation analysis (CCA). Using of CCA statistical method is usual in climatological studies but it has not yet entered into the Iran climatological studies.
Data and Methods:
Two main data sets were used in this study; the time series of 500 hPa monthly relative vorticity fields and monthly precipitation for 97 Iran stations during the rainy season (November to February). The study area extends from latitude of 10 to 70 degrees north and a longitude of 10 to 70 degrees east over a region affecting Iran's precipitation.
Relative vorticity monthly values (1981-2017) using the U and V component wind values obtained from NCEP-DOE reanalysis databases at grid points spaced by 2.5° at the pressure level of 500 hPa was calculated. NCEP-DOE Reanalysis II that is an improved version of the NCEP-NCAR Reanalysis I model that fixed errors and updated parameterizations of physical processes.
The monthly precipitation data (1981-2017) were received from 97 synoptic stations of Iran and were used after standardization. To achieve the purpose of this study, at first the activity centers of vorticity and precipitation were identified by applying S mode principle component analysis (PCA). Then canonical correlation analysis (CCA) was performed on factor scores of these centers. The 30 years base period (1981-2010) was selected to applying CCA method while the years 2010 to 2017 were used as evidence.
This multivariate statistical method was originally developed by Hotelling in 1936 from an interdependence model and first applied in climatology during the 1980s by Barnett and Preisendorfer and also by Nicholls in 1987.
Canonical correlation analysis is a multivariate statistical technique for analyzing internal relations between a set of multiple independent variables (predictors) and a set of multiple dependent variables (predictants).
This method is often used in atmospheric sciences to identify predictors within the datasets. Relationships between variables are highlighted through CCA.
Results and Discussions:
The application of PCA led to 6 factors for the precipitation over Iran and 18 factors for the 500 hPa relative vorticity, accounting for 72% and 80% of the total variance respectively. The first meaningful factor of Iran precipitation is located in northwest of Iran. Second factor is extended from east to south and the fifth factor is in the Caspian Sea southern coasts.
PCA factor scores time series of the each two sets were used for the subsequent CCA.
CCA yielded three physically reasonable relevant pairs of patterns that describe the simultaneous responses of the precipitation field to the relative vorticity changes accounting for 82.5% of the common variance of both fields.
The first CCA pair exhibits a correlation between the precipitation over east of Iran and the 500 hPa relative vorticity changes in the eastern Mediterranean and the Middle East. The second CCA pair reveals a negative correlation between the precipitation over the Caspian Sea southwest coasts and the relative vorticity activities, centered over Eastern Europe. The third CCA pair shows a correlation between the rain of northwest Iran and the relative vorticity activities over Turkey, Cyprus, and the Black sea.
The influence of the relative vorticity centers of activity in different regions of the Middle East, eastern Mediterranean, and Europe that appear in the form of the positive or negative relative vorticity on the Iran precipitation anomalies was investigated.
In this case, three linear combinations were found by the canonical correlation analysis technique.
According to the first pattern, negative vorticity centered over the eastern Mediterranean and the Middle East is associated with below normal precipitation over the east and southern regions of Iran. In fact, the negative vorticity region in the eastern Mediterranean and the Middle East indicates the location of the ridge and the east of Iran locates in front of this ridge that is the area of descending movements that result in weather stability in this part of Iran.
The second pattern reveals that however the relative vorticities will go up to the negative values in Eastern Europe; the rainfall will be higher on the southwest margin of the Caspian Sea. In the winter when Eastern Europe is located in ascending region of western winds (ridge), the Caspian Sea would be located in descending region of mid-level (trough) which will increase weather instability and increase precipitation in the southern coast.
The third pattern says that as far as the relative vorticity tends to positive values in the Cyprus and Turkish cyclogenesis areas, the rainfall will be higher in the northwest of Iran. That’s because the extension of western winds in this area in winter could bring Mediterranean humidity to the northwest and west of Iran with no mountainous barrier.
Therefore, these results can be used to increase rainfall forecasting skills in different parts of the country. The results of this study confirm the results of the study of Rezaei et al. (2012 and 2013).