Meteorol. Appl. 14: 69–78 (2007) Published online in Wiley InterScience (www.interscience.wiley.com) DOI: 10.1002/met.7 Spatial distribution of… [620205]

METEOROLOGICAL APPLICATIONS
Meteorol. Appl. 14: 69–78 (2007)
Published online in Wiley InterScience
(www.interscience.wiley.com) DOI: 10.1002/met.7
Spatial distribution of climatic indices in northern Greece
E. Baltas*
AUTH, Agricultural School, Department of Hydraulics, Soil Sci ence and Agricultural Engineering, 540 06 Thessaloniki, Greecee
ABSTRACT: Results of research work concerning the spatial estimation of precipitation and air temperature and the
representation of different climatic indices, which assess the climate of a region in northern Greece, are reported. The
climatic indices used were the Johansson Continentality Index, the Kerner Oceanity Index, the De Martonne Aridity
Index and the Pinna Combinative Index. Data from 15 meteorological stations located in northern Greece during the
period 1965–1995 were analysed and processed on a monthly b asis and graphs of precipitation and temperature were
constructed for use in agricultural applications. K ¨oppen’s classification was also investigated for the 15 stations. GIS
interpolation techniques, such as the inverse distance wei ghted (IDW) function, were used for the areal estimation of
the above mentioned parameters in northern Greece and an e valuation of the climate indices was made, based on the
results. Copyright 2007 Royal Meteorological Society
KEY WORDS climatic index; climate; areal; GIS; northern Greece
Received 12 July 2006; Revised 20 November 2006; Accepted 20 November 2006
1. Introduction
The present work focuses on the spatial distribution of
various climatic indices tha t determine the climatic con-
ditions of a region. These indices were calculated fromthe processing and analysis of raw air-temperature andprecipitation data of the study period 1965–1995, from
15 meteorological stations (Figure 1, Table I) located in
northern Greece. The climatic indices that were calcu-lated are the Johansson Continentality Index, the KernerOceanity Index, the De Martonne Aridity Index and the
Pinna Combinative Index. The first two belong to the cat-
egory of continental-oceanic indices, while the other twocan be categorized as aridity–humidity indices. The spa-tial representation of the climatic variables and indiceswas performed by the use of GIS techniques. ESRI’s
ArcGIS Desktop-ArcInfo 8.3 software was used for that
purpose. Conclusions were made regarding the effect ofthe country’s geomorphology on the spatial distributionof precipitation and air temperature, as well as regarding
the results of the statistical analysis between the indices
of the same category. Finally, an evaluation of the cli-matic indices was made, suggesting which one is moreappropriate for the study area, based on the results.
The weather conditions and the climate of a region
constitute basic factors for the growth and develop-
ment of plants. Although the correlation between thedevelopmental stages of plants and the meteorologi-cal parameters is empirical, it has been proved very
useful for the forecasting of agricultural production
(Dalezios and Zarpas, 1996; Dalezios et al., 2000). The
* Correspondence to: E. Baltas, AUTH, Agricultural School, Depart-
ment of Hydraulics, Soil Science and Agricultural Engineering, 540 06Thessaloniki, Greece. E-mail: [anonimizat] factors taken into consideration in phenologi-
cal studies are usually climatic , particularly temperature
and precipitation time series (Rosenberg et al., 1983;
Charles-Edwards, 1984). Tempe rature-precipitation dia-
grams were constructed based on the monthly mean val-
ues of precipitation and air temperature of the study
period 1965–1995. Walter (1955); Gaussen (1956) and
Bagnouls and Gaussen (1957) proposed the combined
representation of temperature and precipitation values
through the year to display climatic patterns. Walter(1955) produced the well known standardized format for
graphically representing the Bagnouls and Gaussen cli-
matic diagrams. Their use was successively extended to
the complete world and the Klimadiagramm Weltatlas
by Walter and Lieth (1960) provided what is probably
still the clearest picture of world climatic types. These
temperature-precipitation diagrams became very popular
and have often been applied in vegetation studies to show
the associated climatic features. However, despite the
fact that these diagrams allow markedly different climatic
types to be distinguished by eye, they are not an efficient
way of detecting climatic gradients. This is why indices
continue to be used to define numerical limits between
types, especially in the Mediterranean region (Mazzoleni
et al., 1992).
2. Study area
The study area is the most productive of the coun-
try in terms of economic indices, mainly in the pri-
mary sector (agriculture, livestock farming) and strongly
underpins the national economy of Greece. The pri-
mary sector presents the greatest development capabili-
ties, because the region has rich soil and water resources.
Copyright 2007 Royal Meteorological Society

70 E. BALTAS
LEGEND
0 60 120 180 240KmSEN
W
Latitude: from 38 °58′ to 41 °42′
Longitude: from 19 °40′ to 26 °37′UTM Coordinates
of the Study area:
Figure 1. The sites of the meteorological stati ons in the study area. This figure is available in c olour online at www.interscience.wiley.com/ma
Table I. The meteorological stations and the ir corresponding geographical coordinates
(latitude and longitude) and elevations.
Station Latitude Longitude Elevation above mean
sea level (m)
Alexandroupoli 40 °51/prime13.77/prime/prime25°56/prime45.80/prime/prime’3 . 1 0
Arta 39 °09/prime46.75/prime/prime20°54/prime40.90/prime/prime9.30
Corfu 39 °36/prime31.38/prime/prime19°54/prime40.01/prime/prime1.50
Florina 40 °46/prime54.92/prime/prime21°24/prime15.03/prime/prime692.00
Ioannina 39 °41/prime32.25/prime/prime20°49/prime03.77/prime/prime477.00
Kavala 40 °56/prime05.82/prime/prime24°24/prime22.65/prime/prime5.00
Kozani 40 °17/prime12.33/prime/prime21°50/prime21.53/prime/prime621.00
Larisa 39 °38/prime36.61/prime/prime22°27/prime30.47/prime/prime71.30
Lemnos 39 °55/prime20.49/prime/prime25°14/prime58.19/prime/prime1.70
Methoni 40 °30/prime22°37/prime61.60
Orestiada 41 °30/prime00/prime/prime26°31/prime00/prime/prime43.00
Serres 41 °04/prime27.23/prime/prime23°31/prime36.41/prime/prime32.50
Thessaloniki 40 °32/prime00/prime/prime22°58/prime00/prime/prime4.00
Trikala 39 °31/prime29.82/prime/prime21°45/prime46.35/prime/prime109.00
Vo l o s 3 9 °22/prime00/prime/prime22°57/prime00/prime/prime3.00
The country’s transboundary water resources are located
in this region. The vegetation of the area varies fromgrasslands to dense forest, while the climate is temperate
and humid with substantial seasonal variation in temper-
ature. The geomorphological and climatological condi-
tions are strongly interrelated. The river network and the
topography of the area varies from narrow gorges to wide
flood plains, while the geomorphology of the study area
is characterized by bold relief, with large flat areas.The most intense rainstorms are produced by the
passage of depressions, usually accompanied by coldfronts (and rarely by warm fronts) approaching from the
west, southwest or northwest. Frequent and rapid changes
in weather are caused by frontal air mass activity resulting
in frequent flash floods. A convectional weather type
(characterized by a cold upper air mass that produces
dynamic instability) is also responsible for many intense
storms, especially in the summer (Kurz and Fontana,
Copyright 2007 Royal Meteorological Society Meteorol. Appl. 14: 69–78 (2007)
DOI: 10.1002/met

SPATIAL DISTRIBUTION OF CLIMATIC INDICES IN NORTHERN GREECE 71
2004). The orography of the Pindos mountain range plays
an important role in precipitation and runoff regimes inGreece. The mean annual precipitation exceeds 1500 mmin the mountainous areas of western Greece, whereasin eastern regions of the country it may be as low
as 300 mm. Analysis of precipitation over the years
1965–1995 shows that annual precipitation is decreasing.
3. Data analysis
The raw daily time step data series of precipitation and
air temperature from 15 meteo rological stations located
in northern Greece that constitute the study area wereacquired for the study period 1965–1995. These stationsare operated by the Greek National Meteorological Ser-
vice. Each station’s data were analysed and processed for
the calculation of the mean monthly and annual valuesof precipitation and air temperature. These mean valuesconstituted the base for the calculation of the Johans-son Continentality Index, the Kerner Oceanity Index, theDe Martonne Aridity Index and the Pinna Combinative
Index.
3.1. Johansson Continentality Index
The Johansson Continentality Index is used for the cli-
matic classification between continental and oceanic cli-
mates. The index is calculated by the following formula
(Flocas, 1994; Chronopoulou–Sereli, 1996):
k=1.7E
sinf−20.4 (1)
where:
E=annual range of monthly mean air temperatures,
in°C, (difference between the maximum and minimum
monthly mean air temperatures) and f=station’s geo-
graphical latitude.
The value of the annual amplitude of air temperature
is used to determine the continentality of the climate.
There are many other methods to determine the indexof continentality of a region, but the above formula isthe most often used in many studies (Sj ¨ogersten, 2004;
Filatov et al., 2005). The climate is characterized as
marine when kvaries between 0 and 33, as continental
when kvaries between 34 and 66, and as exceptionally
continental when kvaries between 67 and 100.
3.2. Kerner Oceanity Index
Kerner, motivated by the fact that in marine climates the
spring months are colder than the autumn months, formed
the thermoisodynamic fractio n (Retuerto and Carballeira,
1992; Gavilan, 2005):
k
1=100 (To−Ta)
E(2)
where Toand Taare the October and April mean values of
air temperature respectively and E is the annual range ofmonthly mean air temperatures, in °C. Small or negative
values of k1imply a continental climate, while larger ones
imply oceanity (Zambakas, 1992). More specifically, in
the present study, when the Kerner Oceanity Index is
higher than 10 the climate is characterized as oceanic.
3.3. De Martonne Aridity Index
This is an aridity–humidity index and is only applica-
ble locally. Aridity is the degree to which a climatelacks effective, life-promoting moisture; the opposite ofhumidity, in the climate sense of the term (AmericanMeteorological Society, 2006). A measure of aridity of
a region, proposed by De Martonne (1925), is given by
the following relationship:
I
DM=P
T+10(3)
where Pis the annual mean precipitation (mm) and T
(°C) the annual mean air temperature.
An increase in the value of IDM, at constant temper-
ature, implies an increase of precipitation. The De Mar-
tonne index climatic classification based on the values of
IDMand Pi ss h o w ni nT a b l eI I .
The monthly value of the De Martonne Aridity Index
is calculated by the following equation:
Im=12P/prime
T/prime+10(4)
where P/primeand T/primeare the monthly mean values of
precipitation and air temperature for the consideredmonth. When the value of I
mis lower than 20 then
the land in this month needs to be irrigated (Zambakas,
1992).
3.4. Pinna Combinative Index
Pinna developed the combinative index Ip(Zambakas,
1992):
Ip=1
2/parenleftbiggP
T+10+12P/prime
d
T/prime
d+10/parenrightbigg
(5)
where Pand Tare the annual mean values of precipita-
tion and air temperature and P/prime
d,T/prime
dare the mean values of
precipitation and air temperature of the driest month. Thisindex describes in a better way the regions and seasons,
Table II. De Martonne index climatic classification.
Climate Values of IDM Values of P(mm)
Dry IDM <10 P< 200
Semi-dry 10 ≤IDM≤20 200 ≤P< 400
Mediterranean 20 ≤IDM <24 400 ≤P< 500
Semi-humid 24 ≤IDM <28 500 ≤P< 600
Humid 28 ≤IDM <35 600 ≤P< 700
Very humid a.35 ≤IDM≤55 700 ≤P< 800
b.IDM >55 P> 800
Copyright 2007 Royal Meteorological Society Meteorol. Appl. 14: 69–78 (2007)
DOI: 10.1002/met

72 E. BALTAS
where irrigation is necessary since it takes into account
the precipitation and air temperature of the driest month.
When the value of Ipis less than 10 ( Ip<10) the climate
is characterized as dry and when the value of Ipvaries
between 10 and 20 (10≤Ip≤20)the climate is consid-
ered semi-dry Mediterranean with formal Mediterranean
vegetation.
3.5. Areal estimation of t he climatic variables and
indices
The maps depicting the spatial distribution of the annual
mean precipitation, the annual mean air-temperature, the
Johansson index, the Kerner index, the De Martonneindex and the Pinna index were developed using ESRI’s
software ArcGIS Desktop-ArcInfo 8.3. After the calcu-
lation of the aforementioned variables and indices, the
spatial integration of the stations’ point values followed
and the raster coverages of the variables and indices were
developed via the ‘Interpola tion to Raster’ Inverse dis-
tance weighted (IDW) function, included in the SpatialAnalyst extension of ArcMap. Interpolation predicts val-
ues for cells in a raster from a limited number of sample
data points and can be used to predict unknown values
for any geographic point data: for example elevation or
precipitation. The IDW method estimates cell values byaveraging the values of sample data points in the vicin-
ity of each cell. The closer a point is to the centre of the
cell being estimated, the more influence, or weight, it has
in the averaging process. Th is method assumes that the
variable being mapped decreases in influence with dis-
tance from its sampled location (McCoy and Johnston,
2001).
The parameters of the IDW method are the power
and search radius types. By defining a high power,
more emphasis is placed on the nearest points, and the
resulting surface will have more detail (i.e. less smooth).
Specifying a lower power will give more influence to
the points that are further away, resulting in a smoother
surface. A power of 2 is most commonly used as thedefault, as in this case. IDW also has two options: fixed
search radius type and variable search radius type. In
the present case, the variable search radius type was
selected. With a variable search radius, the number of
points used in calculating the value of the interpolatedcell is specified, which makes the radius distance vary
for each interpolated cell, depending on how far it has
to search around each interpolated cell to reach the
specified number of input points (ArcMap Help Menu).
The specified number of points was 12 and no maximum
distance was defined. Finally, the output cell size was
set to 1000 m. The above parameters were used for theproduction of each raster.
Finally, the function CONTOUR included in the ‘Sur-
face Analysis’ toolset of the spatial analyst extension was
used for the construction of the final contour maps of each
variable and index.3.6. Temperature–precipitation diagram s
The estimated monthly mean values of precipitation and
air temperature were necessary for the construction ofthe temperature-precipitation diagram of each station.Gaussen (1956) defines a month as dry, when:
P
/prime<2T/prime(6)
where P/primeand T/primeare the monthly mean values of
precipitation (mm) and air temperature ( °C), respectively.
On the basis of the fact that the extent of biological
drought depends on the humidity of the air, Gaussensupplemented the above equation and characterized amonth as dry, when the precipitation volume is:
1. less than 10 mm and the mean monthly value of
temperature is less than 10
°C;
2. less than 25 mm and the mean monthly value of
temperature varies between 10 and 20 °C;
3. less than 50 mm and the mean monthly value of
temperature varies between 20 and 30 °C;
4. less than 75 mm and the mean monthly value of
temperature is greater than 30 °C.
4. Results
Figure 2 depicts the mean annual areal precipitation
for 30 years (1965–1995). One can observe the highvariability of precipitation varying from 1150 mm in thewestern part of the study area to 450 in the eastern part.This is attributed to the Pindos mountain range, whichinterrupts the eastern movement of weather systems,dividing the country into windward, high precipitation,western areas and leeward, low precipitation, easternareas. The western air masses ascend the windward sideof the Pindos mountains, cool and lose most of theirmoisture. When they descend the other side they arewarmed by adiabatic sinking.
The mean annual areal temperature for the study
area, derived after the analysis and processing of thedata for the study period, is depicted in Figure 3. It isobvious that the spatial distribution of the mean annualair temperature is not uniform. It is affected by manyfactors, the most important of which is the direction of themountain range parallel to the direction of the coastline(Zambakas, 1992). The mountainous terrain contributesto the decrease in air temperature from the coast to theinterior (e.g. the region of Florina). Additionally, duringthe winter months the mountain range of Pindos protectsthe western areas of Greece from the northeastern cold airmasses and that, in conjunction with the areas’ proximity
to the sea, contributes to their greater mean temperatures.
Other factors affecting the spatial distribution of airtemperature are the wind direction local conditions suchas the urbanization of a region (e.g. Thessaloniki) whichformulates a special microclimate in the region. In flatareas, such as Thessaly, which is located at the southernpart of the study area, the temperature increases from
Copyright 2007 Royal Meteorological Society Meteorol. Appl. 14: 69–78 (2007)
DOI: 10.1002/met

SPATIAL DISTRIBUTION OF CLIMATIC INDICES IN NORTHERN GREECE 73
LEGEND
416 – 500
501 – 600601 – 700
701 – 800801 – 900
901 – 10001001 – 1100Study period: 1965 – 1995
0 60 120 180 240KmSEN
WMean annual areal
precipitation (mm)
Figure 2. The mean annual areal precipitation (mm) of the study area. This figure is available in col our online at www.interscience.wiley.com/ma
LEGEND
12 – 13
14
1516
17
18Study period: 1965 – 1995
0 60 120 180 240KmSEN
WMean annual areal air
temperature ( °C)
Figure 3. The mean annual areal air temperature ( °C) of the study area. This figure is available in c olour online at www.interscience.wiley.com/ma
Copyright 2007 Royal Meteorological Society Meteorol. Appl. 14: 69–78 (2007)
DOI: 10.1002/met

74 E. BALTAS
LEGEND
19 – 20
21 – 2223 – 2425 – 2627 – 2829 – 30
31- 32
33- 34
35- 36
37- 38 0 60 120 180 240KmSEN
WJohansson Continentality
IndexStudy period: 1965 – 1995
Figure 4. The Johansson Continentality Index i n the study area. This figure is available in col our online at www.interscience.wiley.com/ma
LEGEND
0 – 3
4 – 6
7 – 910 – 12
13 – 1516 – 18
19 – 21
22 – 2425 – 27
28 – 30 0 60 120 180 240KmSEN
WKerner Oceanity Index
Study period: 1965 – 1995
Figure 5. The Kerner Oceanity Index in the st udy area. This figure is available in colour on line at www.intersci ence.wiley.com/ma
Copyright 2007 Royal Meteorological Society Meteorol. Appl. 14: 69–78 (2007)
DOI: 10.1002/met

SPATIAL DISTRIBUTION OF CLIMATIC INDICES IN NORTHERN GREECE 75
the coast to the interior due to the cooling effect of the
coastal air masses.
The spatial distribution of the Johansson Continentality
Index is depicted in Figure 4. Its values vary between19 and 38 for the entire area of study. At seven out
of the 15 stations the index value was higher than 33
(limit), denoting continental climate: six out of these
seven stations were located in the interior, away from
the sea. At the rest of the stations the climate wascharacterized as marine.
Figure 5 depicts the Kerner Oceanity Index, which
varies between 0.47 (Florina) and 28.74 (Methoni). Thespatial distribution of Kerner’s index is similar to that
of Johansson. The lower values, implying a continental
climate, are distributed to the stations located in the
interior of the country. More specifically, at eight out of
the 15 stations the climate was defined as continental.The classification at seven out of these eight stations
was the same with Johansson’ s classification, denoting
an agreement between the results of the two methods.
A correlation (Figure 6) was also performed between
the values of Johansson’s and Kerner’s indices. The highcorrelation coefficient proved the above observed agree-
ment between the estimations. The statistical analysis
indicated that the Johansson index limit value of 33 coin-cides with the Kerner index value of 10, meaning that
when the Kerner Oceanity Index is higher than 10 the
climate is characterized as oceanic. Evaluating the results
of the two indices, the Johansson index is preferable to
the Kerner index for the definition of a station’s climateas continental or oceanic. The basic reason is that there35
30
25
20
15Kerner Oceanity Index10
5
0
R2 = 0.9011LEGEND
Linear0
51 0 1 5 2 0
Johansson Continentality Index25 30 35 40
Figure 6. Statistical analysis between the indices of Johansson and
Kerner. This figure is available in colour online at www.interscience.
wiley.com/ma
are distinct limit values between the climate categories
of the Johansson classification. On the contrary, the limit
value separating continental from oceanic climate in the
Kerner classification is not known and was determined
in the present study in the statistical analysis.
Figure 7 depicts the De Martonne Aridity Index, whose
values range from 16.28 (Larissa), denoting a semi-
dry climate, to 43.01 (Ioannina), denoting a very humid
climate. The values of the index increase gradually from
the eastern to the western coast, covering almost the
entire range of the classificati on’s climate categories. This
is in line with the spatial distribution of precipitation,
which increases gradually from east to west.
LEGEND
16 – 18
19 – 21
22 – 24
25 – 27
28 – 3031 – 33
34 – 36
37 – 39
40 – 42
43 – 45 0 60 120 180 240KmSEN
WDe Martonne Aridity Index
Study period: 1965 – 1995
Figure 7. The De Martonne Aridity Index in the study area. This figure is available in colour online at www.intersci ence.wiley.com/ma
Copyright 2007 Royal Meteorological Society Meteorol. Appl. 14: 69–78 (2007)
DOI: 10.1002/met

76 E. BALTAS
LEGEND
11 – 12
13 – 1415 – 16
17 – 1819 – 20
21 – 22
23 – 24
25 – 26
27 – 28 0 60 120 180 240KmSEN
WPinna Combinative Index
Study period: 1965 – 1995
Figure 8. The Pinna Combinative Index in the study area. This figure is available in colour online at www.intersci ence.wiley.com/ma
The Pinna Combinative Index is depicted in Figure 8.
At 12 out of the 15 stations, the values of the index ranged
from 10 to 20, implying a semidry Mediterranean climatewith formal Mediterranean vegetation. At the remainingthree stations (Arta, Ioannina, Corfu) the values of the
index were higher than 20. These three stations are
located at the northwestern side of Greece and theannual precipitation volume ( >1000 mm) at each of these
stations is higher than that of any station in the study area.
Furthermore, the spatial distribution of the Pinna index issimilar to that of De Martonne index. The maps generally
show that the climate gets drier from the western to the
eastern coast.
A correlation (Figure 9) was also performed between
the values of De Martonne and Pinna indices. A high
correlation coefficient was noticed verifying the similarspatial distribution of the two indices. Evaluating the
results of the aridity-humidity indices of De Martonne
and Pinna, the De Martonne Index is more appropriatefor the study area, since it defines more preciselythe climate of each station. Its classification consists
of six climate categories, ranging from dry to very
humid, instead of just two categories of the Pinnaindex.
Figure 10 depicts the monthly aridity index of De
Martonne for August, a month with very low precipitationvolume and high water demands for agricultural use.
The values of the index are under 20, implying that
the entire study area needs irrigation. Furthermore, the30
25
20
15Pinna Combinative Index10
5
0
R2 = 0.8865LEGEND
Linear0
51 0 1 5 2 0
De Martonne Aridity Index25 30 35 40 45 50
Figure 9. Statistical analysis between the indices of De Mar-
tonne and Pinna. This figure is available in colour online at
www.interscience.wiley.com/ma
spatial distribution of the index emphasizes increased
irrigation requirement on the eastern flat plains of the
Thessaly region, located in the southern part of the study
area.
The bar diagrams of precipitation and air temper-
ature, as well as the temperature-precipitation dia-grams of the 15 meteorological stations, were derived.These diagrams depict the monthly mean values ofprecipitation and air temperature for the study period
1965–1995. Figures 11 and 12 show the bar diagrams
for temperature and precipitation, respectively, for theTrikala meteorological station. Figure 13 depicts the
Copyright 2007 Royal Meteorological Society Meteorol. Appl. 14: 69–78 (2007)
DOI: 10.1002/met

SPATIAL DISTRIBUTION OF CLIMATIC INDICES IN NORTHERN GREECE 77
LEGEND
1 – 2
3 – 4
5 – 67 – 8
9 – 10
11 – 12
13 – 14 0 60 120 180 240KmSEN
WMonthly aridity index of De
Martonne (August)Study period: 1965 – 1995
Figure 10. The monthly aridity index of De Martonne in the st udy area (August). This figure is available in colour online at
www.interscience.wiley.com/ma
temperature-precipitation diagram of the same station.
Five of the months (from May to September) at Trikalastation are characterized as dry according to Gaussen
classification.
Climate was also classified according to K ¨oppen’s
methodology (http://clem.mscd.edu/ ∼wagnerri/Clima-
tology/classification.htm, K ¨oppen’s Climate Classifica-
tion, 2006). It was found that at three of the stations
(Florina, Kozani, Ioannina) the climate type is Cfa (longand hot summer and precipitation in all seasons) and atthe remaining stations the climate type is Csa (Mediter-
ranean climate – hot and dry summer with mild winter).
30
25
2015Air Temperature ( °C)
10
5
0
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
Figure 11. The monthly mean air temperature for the Trikala sta-
tion. This figure is available in colour online at www.interscience.
wiley.com/ma
30405060708090100
20Precipitation (mm)
10
0
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
Figure 12. The monthly mean precipitation for the Trikala station. This
figure is available in colour online at www.interscience.wiley.com/ma
5. Conclusions
This research work has shown that the north–south direc-
tion of the Pindos mountain range, which is parallel to
the coastline, constitutes a cruc ial factor in the spatial dis-
tribution of precipitation and air temperature. Extending
vertical to the western air masses, the mountain range
divides the country into western high and eastern lowprecipitation areas.
The classification of the climate of the 15 stations
into continental and oceanic, according to the Johansson
Continentality Index was in full agreement with the
classification based on Kerner Oceanity Index. Besides,
the statistical analysis of the two indices’ values indicated
a high correlation coefficient, verifying their similar
Copyright 2007 Royal Meteorological Society Meteorol. Appl. 14: 69–78 (2007)
DOI: 10.1002/met

78 E. BALTAS
3035404550
Air Temperature
25
20
15Air Temperature ( °C)
10
5
0 0102030405060708090100
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
Precipitation (mm)Precipitation
Figure 13. The temperature – precipitation diagr am of the Trikala station. The measurement unit of precipitation (on the right vertical axis) in
mm is double the temperature’s measurement unit (on the left vertical axis) in °C. The twelve months are marked on the horizontal axis. This
figure is available in colour online at www.interscience.wiley.com/ma
spatial distribution. The J ohansson index is preferable
to the Kerner index owing to the former’s distinct limitvalues separating continental from oceanic climates.
The next classification of the stations’ climate, accord-
ing to the De Martonne Aridity Index and the PinnaCombinative Index, also resulted in the agreement of thetwo indices’ results. Their spa tial distribution was similar
and their correlation coefficients were high. Evaluating
the results of these two methods, the De Martonne indexled to a more precise definition of each station’s climatedue to its more climate categories, in contrast to the few
climate categories of Pinna index.
The temperature-precipitation diagrams allow the dis-
tinction between markedly different climatic types byeye, but they do not define climate gradients. K ¨oppen’s
classification was also performed for the 15 stations and
it was found that twelve of them have climate type Csaand three of them Cfa.
References
American Meteorological Society. 2006. Glossary of Meteorology,
http://amsglossary.allenpress.
com/glossary, accessed 10 May 2006.
Bagnouls F, Gaussen H. 1957. Les climats biologiques et leur
classification. Annales De Geographie 355: 193–220.
Charles-Edwards DA. 1984. Physiological Determination of Crop
Growth . Academic Press: Sydney, Australia; pp 161.
Chronopoulou – Sereli A. 1996. Courses of Agricultural Meteorology .
Publications Agricultural University of Athens; Athens, OH.
Dalezios NR, Zarpas CD. 1996. The role of NDVI and meteorological
indices in the phenology of cotton, vineyards and apples in
Greece. International Symposium on A pplied Agrometeorology and
Agroclimatology, 24–26 April 1996 (Luxembourg: Office for Official
Publications of the European Communities), pp. 185–190.
Dalezios N, Domenikiotis X, Tzortzios S, Loukas A, Tzanetopoulou I.
2000. The importance of the agrometeorological indicators on thedevelopment of the wheat crop in the area of Larisa. In Practical of
the 2nd National Congress of Agricultural Mechanics , V olos.
De Martonne E. 1925. Trait´ ed eG ´ eographie Physique : 3 tomes, Paris.
Filatov N, Salo Y, Nazarova L. 2005. Effect of climate variability
on natural water bodies in Northwest Russia. 15th International
Northern Research Basins Symposium and Workshop .L u l e ˚at o
Kvikkjokk: Sweden.
Flocas AA. 1994. Courses of Meteorology and Climatology . Ziti
Publications: Thessaloniki.
Gaussen H. 1956. Le XVIII congres inte rnational de Geographie, Rio
de Janeiro, Aout 1956. Annales De Geographie 353: 1–19.
Gavilan RG. 2005. The use of climatic parameters and indices in
vegetation distribution. A case st udy in the Spanish Sistema Central.
International Journal of Biometeorology 50: 111–120.
K¨oppen’s Climate Classification. 2006. Website: http://clem.mscd.edu/
∼wagnerri//Climatology/classifica tion.htm, (Date accessed: 10/05/
2006).
Kurz M, Fontana AD. 2004. A case of cyclogenesis over the
western Mediterranean Sea with extr aordinary convective activity.
Meteorological Applications .11: 97–113.
Mazzoleni S, Lo Porto A, Blasi C. 1992. Multivariate Analysis of
Climatic Patterns of the Mediterranean Basin ,Vegetatio 98: 1–12,
1992. 1992 . Kluwer Academic Publishers, Journal of Plant
Ecology, Publisher: Springer Netherlands.
McCoy J, Johnston K. 2001. Using ArcGIS Spatial Analyst .G I Sb y
ESRI: ESRI, Redlands, CA, USA.
Retuerto R, Carballeira A. 1992. Use of direct gradient analysis to
study the climate-vegetation rel ationships in Galicia, Spain. Vegetatio
101: 183–194, 1992 Kluwer Academic Publishers, Printed in
Belgium.
Rosenberg NJ, Bland BL, Verma SB. 1983. Microclimate: The
Biological Environment .
Sj¨ogersten S, Wookey PA. 2004. Decomposition of mountain birch leaf
litter at the forest-tundra ecotone in the Fennoscandian mountains in
relation to climate and soil conditions. Plant and Soil 262: 215–227,
2004 Kluwer Academic Publishers. Printed in the Netherlands.
Walter H. 1955. Die Klima-Diagramme als Mittel zur Beurteilung
der Klimaverhtltnisse fur kologi sche, vegetationskundliche und
landwirtschaftliche Zwecke. Berichte Der Deutschen Botanischen
Gesellschaft 68: 321–344.
Walter H, Lieth H. 1960. Klimadiagramm-weltatlas . Gustav Fisher
Verlag: Vienna.
Zambakas J. 1992. General Climatology . Department of Geology,
National & Kapodistrian University of Athens: Athens, Greece.
Copyright 2007 Royal Meteorological Society Meteorol. Appl. 14: 69–78 (2007)
DOI: 10.1002/met