184
Murat TUrke~
the homogeneity test, and checked by the station files. These station history files were
prepared especially for the studies of precipitation and temperature, and include
necessary information, such as changes of meteorological instruments and of height
and location of the stations for about 80 stations, in order to detect inhomogeneity in
the series. This type of information is useful to decide whether a jump arose from a
relocation of the station, or it happened due to a low-frequency fluctuation (or a
strong persistence) in the series. For instance, by using additional information from
the station history files, we found that significant inhomogeneities in the annual and
seasonal air temperature series of Antalya, Anamur and Fethiye stations had occurred
owing to relocation of the stations.
However, it was seen that monthly series could not be a good indicator for some
inhomogeneity types, such as in the series that were exposed to a location change
without considerable change in station height. This situation also happens when a
station relocation occurs from a relatively inland site to a new site on coast or on
near-coast. Effect of the local climatic conditions of a station could be different in
course of the year, when a relocation of the station occurs. Thus, to make a best
decision on whether a climatic series is appropriate for the time-series analysis or not,
information from both objective and subjective analyses for the seasonal series should
be examined.
According to the results of the objective and subjective homogeneity analyses,
annual and seasonal precipitation series of 91 stations, and mean annual temperature
and aridity index series of 55 stations were found to be homogeneous, with respect
to abrupt changes in the series (Tiirke~, 1995, 1996a and 1999). Regarding the timeseries analyses, this study uses the precipitation series of these 91 stations.
Geographical distribution of the 91 stations over the rainfall regime regions of Turkey
is shown in Fig. I. The rainfall regime regions of Turkey were originally designated
by taking into account the seasonality of Turkish precipitation and the physical
geographical conditions of Turkey. The delimitation of these regions and the
description of their rainfall characteristics were widely discussed and presented by
Tiirke~ (l996a, 1998a).
The rainfall regime regions of Turkey and their fundamental characteristics are
defined as follows:
I. Black Sea (BLS): Uniform rainy with a maximum in autumn; temperate.
2. Marmara Transition (MRT): Quite uniform rainy with a warm and light rainy
summer.
3. Mediterranean (MED): Markedly seasonal with a cool and heavy rainy winter and
a hot dry summer; humid and semi-humid subtropical.
4. Continental Mediterranean (CMED): Seasonal with a rainy winter and spring and
a severe hot dry summer; semi-arid and dry semi-humid subtropical.
5. Mediterranean to Central Anatolia Transition (MEDT): Moderate rainy winter and
spring.
6. Continental Central Anatolia (CCAN): Cool rainy spring and cold rainy winter, and
warm and light rainy summer; semi-arid and dry semi-humid steppe.
7. Continental Eastern Anatolia (CEAN): Cool rainy spring and early summer with
a very cold and snowy winter; dry semi-humid and semi-humid steppe and highland.
Additional eight and thirty-five stations having relatively shorter series of
observations were included to this main 91 and 55 stations, respectively, for
Murat TUrke~
the homogeneity test, and checked by the station files. These station history files were
prepared especially for the studies of precipitation and temperature, and include
necessary information, such as changes of meteorological instruments and of height
and location of the stations for about 80 stations, in order to detect inhomogeneity in
the series. This type of information is useful to decide whether a jump arose from a
relocation of the station, or it happened due to a low-frequency fluctuation (or a
strong persistence) in the series. For instance, by using additional information from
the station history files, we found that significant inhomogeneities in the annual and
seasonal air temperature series of Antalya, Anamur and Fethiye stations had occurred
owing to relocation of the stations.
However, it was seen that monthly series could not be a good indicator for some
inhomogeneity types, such as in the series that were exposed to a location change
without considerable change in station height. This situation also happens when a
station relocation occurs from a relatively inland site to a new site on coast or on
near-coast. Effect of the local climatic conditions of a station could be different in
course of the year, when a relocation of the station occurs. Thus, to make a best
decision on whether a climatic series is appropriate for the time-series analysis or not,
information from both objective and subjective analyses for the seasonal series should
be examined.
According to the results of the objective and subjective homogeneity analyses,
annual and seasonal precipitation series of 91 stations, and mean annual temperature
and aridity index series of 55 stations were found to be homogeneous, with respect
to abrupt changes in the series (Tiirke~, 1995, 1996a and 1999). Regarding the timeseries analyses, this study uses the precipitation series of these 91 stations.
Geographical distribution of the 91 stations over the rainfall regime regions of Turkey
is shown in Fig. I. The rainfall regime regions of Turkey were originally designated
by taking into account the seasonality of Turkish precipitation and the physical
geographical conditions of Turkey. The delimitation of these regions and the
description of their rainfall characteristics were widely discussed and presented by
Tiirke~ (l996a, 1998a).
The rainfall regime regions of Turkey and their fundamental characteristics are
defined as follows:
I. Black Sea (BLS): Uniform rainy with a maximum in autumn; temperate.
2. Marmara Transition (MRT): Quite uniform rainy with a warm and light rainy
summer.
3. Mediterranean (MED): Markedly seasonal with a cool and heavy rainy winter and
a hot dry summer; humid and semi-humid subtropical.
4. Continental Mediterranean (CMED): Seasonal with a rainy winter and spring and
a severe hot dry summer; semi-arid and dry semi-humid subtropical.
5. Mediterranean to Central Anatolia Transition (MEDT): Moderate rainy winter and
spring.
6. Continental Central Anatolia (CCAN): Cool rainy spring and cold rainy winter, and
warm and light rainy summer; semi-arid and dry semi-humid steppe.
7. Continental Eastern Anatolia (CEAN): Cool rainy spring and early summer with
a very cold and snowy winter; dry semi-humid and semi-humid steppe and highland.
Additional eight and thirty-five stations having relatively shorter series of
observations were included to this main 91 and 55 stations, respectively, for