نویسندگان
چکیده
کلیدواژهها
عنوان مقاله [English]
To identify temporal and spatial precipitation models in the southern coasts of Caspian Sea, principle component analysis (PCA) and clustering analysis (CA) were used. In this study, using daily precipitation data, measure of central tendency, statistical dispersion and percentages of three variables of duration, amount and intensity of precipitation events in 46 synoptic, climatic and rain gauge stations from 1957 to 2004 were calculated. Southern coasts of Caspian Sea were regionalized on the basis of standardized data of the 3 variables and its regimes were calculated. Precipitation events of each station was divided into four groups including light, moderate, heavy and super heavy and its regimes were explained. The results showed that there are six monotonous precipitation regions in this area as the spatial variations are more in central parts than east and west parts. The mean of rain duration, variance of precipitation duration and precipitation amount as well as skewness and kurtosis of precipitation amount and intensity are more in central and western parts than eastern and mountainous areas. Despite the fact that precipitation mean is maximum from September to December in different regions but its frequency is different. Frequency regimes analysis showed that coastal areas (regions of 1, 2 and 3), mountainous area in west of Guilan Province (part of region of 3) has maximum frequency from September to December. The mountainous areas in center of Mazandaran province (region of 4) as well as the other regions in mountainous areas (region of 5) have the highest frequency of precipitation events from April to August and in winter respectively. Also, the results show that July and summer have minimum frequency of events in all precipitation groups. The light, moderate and heavy precipitation events have maximum frequency events in March and winter as well as the super heavy precipitation events has the highest frequency in October and autumn.
کلیدواژهها [English]