Frequency and Intensity of Sultriness in Istanbul

Serdar Bahad, M Zeki Karagülle


Istanbul Universitesi, Istanbul Tıp Fakültesi

Tibbi Ekoloji ve Hidroklimatolojı ABD

Millet Cad. 126, 34390 Istanbul, Turkey 

Fax : +90 212 531 89 04

e-mail  :

Abstract: The combination of high air temperature and high humidity, calm or light wind and intensive radiation facilitate sultriness. The milieu is perceived as very hot or sultry by human beings depending on above meteorological conditions. This explains the appearance of heat stress affecting human beings, especially in highly sensitive individuals. Our purpose was to obtain the frequency and intensity of sultriness if it were to occur in Istanbul. Sultry index (S) was computed from hourly meteorological data, which were observed at Atatürk airport station. Based on the empirical boundary value for sultriness, i.e. S=30.0, all sultry hours were analysed from May to October for the years 1980-1989. Sultriness was most frequent between 12.00 h. and 15.00 h. in local time, and most intense during daytime especially in July and August. There were great differences in monthly and yearly number of sultry hours. These were related with the frequencies of tropical air masses affecting Istanbul

Key words:  Sultriness ∙Sultry index ∙Heat stress ∙Human comfort ∙Istanbul 

Introduction: Human body perceives the ambient temperature in terms of the heat loss due to conduction, convection, radiation and evaporation. The apparent temperature depends on meteorological factors such as air temperature, mean radiant temperature, humidity and wind speed which all play an important role on heat loss. The heat loss is essentially regulated by the conditions of the surrounding atmosphere. Increase in air temperature, mean radiant temperature, and humidity and decreased wind speed all play a prominent role in the problems of sultriness. The human body cannot lose heat easily by sweating in sultry conditions because of the reduction of evaporation from the body surface.

Sultry weather is always a problem on human health in tropical and sub-tropical regions, especially in the summer months. Human efficiency declines, both physically and intellectually,  when the weather is perceived as sultry.

 Much work has been devoted to study the human comfort, and a large number of bio-meteorological indexes based on different combinations of the above-mentioned variables have been introduced in this respect. The perfect one of these combinations is the Predicted Mean Vote (PMV) derived from human heat balance by Fanger (1982) and discussed in detail by Jendritsky (1990), Höppe (1993, 1999) and Matzarakis (1997, 1999). The application of PMV is rather more difficult than the others because the radiation observations are not generally available in practice.

In the literature, the indexes most widely used to study the sultriness were the effective temperature (ET), water vapour pressure (℮) and equivalent temperature (Teq). The frequencies of sultriness were also determined with the empirical boundary values ET =24.0 ºC,  ℮ =18.8 hPa and Teq =56.0 ºC between comfort and sultry regions (Mayer 1975,1977;Mayer and Abele 1977;Dieterichs 1958,1980); Olaniran (1982). 

King (1955) had defined an empirical sultry index by using Teq, cooling power and atmospheric counter radiation. Leistner (1964) had also presented the relationship between wind speed and Teq. 

Dieterichs (1980) had studied the sultriness by using sultry index (S) considering wind speed and cloudiness in addition to air temperature and humidity. The frequency of sultriness had also been determined with the empirical boundaries ℮= 18.8 hPa and S= 25.0 in his study. 

Methods: Sultry index (S) can be easily calculated by using the equation (1) that was given by Dieterichs (1980). 

S = Td + 0.5 (T + N) – 0.15 [v (35–T)] ½                      (1)                                              

S:    Index value (dimensionless)

Td:  Dew point temperature (˚C)

T:    Dry bulb (air) temperature (˚C)

N:   Total cloudiness (0-8 in eights)

v :   Wind speed  (knots)


The sultry index equation (1) considers wind-chill and atmospheric counter radiation effects due to wind speed and cloudiness.  If N=0 and v=0 in Eq.(1), the curves of  S and Teq are closely similar in the temperature-humidity diagram. For the sake of greater reliability, S is more preferable than Teq, despite the fact that both are found to be approximately similar.  Under these conditions, sultry index value S=30.0 is a suitable value compared to the boundary value Teq =56.0 ˚C for sultriness. Therefore in this study, sultriness had been investigated by using sultry index Eq. (1) and S≥ 30.0 is considered ‛sultry ’ while S< 30.0 ‛not sultry’.

On hourly basis, S was computed from meteorological data, which were recorded between May and October for the years 1980-1989 at the Atatürk airport meteorological station. The station is located at the southwest coastal area nearby Istanbul city center (h=19 m. above sea level, latitude 40º 58′ and longitude 28º 49′ ). 

To determine the frequency of sultriness,  all sultry hours with S ≥ 30.0 were counted.

Hourly and monthly numbers of sultry hours were recorded and the intensity of sultriness was analysed in detail by using the classified index values. Meteorological parameters accompanied by sultriness were also examined. 

Results: Table 1 indicates that sultriness with S ≥30.0 never occurred in the months of May and October for the years 1980-1989. The numbers of sultry hours varied significantly from year to year. The last four-years (1986-1989) were significant because the cumultative hours of sultriness were calculated to be 2320 hours (78.9%) out of a total of 2940 sultry hours for ten years.  Monthly sultry hours with S ≥ 30.0 were most frequent in August with 1255 hours (42.7 %) and in July with 1202 hours (40.9%) out of a total of  2940 sultry hours. Tropical air masses affecting Istanbul occurred almost regularly in July and August, but irregularly in June and September. Therefore, the sultriness was found to be infrequent in June with 343 sultry hours (11.7 %) and especially in September with 140 sultry hours (4.8 %) out of a total of 2940 sultry hours for ten years (Table 1).  The sultry hours with S ≥ 30.0 displayed dispersion during the months of June, July and August but only at the first ten days of September for the years 1980-1989.  The rates of sultry hours to the total time showed that sultriness occurred most often in July (16.13%) and August (16.87%), occasionally in June (4.74%) and September (1.94%) (Table 2).  

Hourly frequencies of sultriness with S ≥30.0 were also analysed for ten years. In Figure 1,  sultriness for all months occurred frequently during the daytime between 0800 h. and 1900 h. and infrequently between 2000 h. and 0700 h. in local time. Especially at night in June and September,  relatively few sultry hours with S ≥ 30.0 were recorded for the years 1980-1989, as depicted in Figure 1.                                            

In Table 3, monthly numbers of sultry hours were categorized according to various levels of  classified  S values for ten years. Intensive sultriness with the value of S ≥ 34.0 occurred in July and August with few sultry hours in June but none in September. The hourly numbers of sultry hours with classified S values are depicted in Figure 2.  It  showed that all of the intensive sultriness ( S ≥ 34.0) occurred between 0800h. and 2300h. in local time, especially during daytime at high air temperatures.

 The important meteorological parameters accompanied by sultriness were examined in detail and  summarized in Table 4. The data indirectly suggested that the sultriness observed in Istanbul was mainly dependent on a higher values of water vapour pressure, i.e. ℮≥18.8hPa. The activity of the cumulus clouds was increased by convection and humid air due to intensive solar radiation especially at noon. Therefore, the sultriness could often occur even if the sky were partly covered by cumulus clouds at low level, although the maximum frequency of sultry hours with S ≥30.0 was in clear sky conditions.

 Total number of sultry hours with S≥ 30.0 was reduced approximately half by wind chill effect due to wind speed in Equation 1. In other words, sultry index values closed to 30.0 were eliminated by wind chill, especially in air temperature less than 25.0 ºC. But  sultriness could occur even in windy (v≥10 knots) conditions at higher temperatures (Table 4). 

Discussion: The results of this study are valid for shady areas by appropriating air temperature as  equal to mean radiant temperature (Tmrt). On the other hand, the meteorological data in this study are related to rural site of Istanbul. Asphalt and buildings in the urban cause an increase in the value of Tmrt by long-wave irradiation, compared to the rural site. The wind chill effect due to wind speed is also decreased by irregular, building structure in the urban. The number of sultry hours in the urban must be more than that observed in this  study because of the increased heat load on human beings.

 Dieterichs (1980) has found only 190 sultry hours with S≥30.0 at North Sea coast in Germany for the years 1966-1976. During the ten-year period (1980-1989), 2940 sultry hours with S≥30.0 had been realized in Istanbul. This great difference in the number of sultry hours between Istanbul and northern coast of Germany shows that the sultriness increases toward the low latitudes. Harlfinger (1975) has indicated that the sultriness chanced in accordance with altitude and latitude. In both studies, sultry index S showed  similar diurnal variation.

 The frequency of tropical air masses affecting Istanbul is very high especially in the summer, but this can vary year by year due to synoptic pattern related to the vicinity of Istanbul. Because of this, frequency of sultriness varies year by year with great differences.  

The sultry period (June-September) included 830 summer days with Tmax ≥ 25.0 ºC and 142 tropical days with Tmax ≥ 30.0 ºC. This can be taken as indicators of the high air temperatures realized in Istanbul. There are two big water masses such as the Black Sea and Marmara Sea, which are situated in the north and south of Istanbul respectively. In high sun season, the intensive solar radiation causes humid air by evaporation over the seas and a heated atmosphere by irradiation over the artificial structure in the city center of Istanbul. The humid air occurring over the Black Sea and Marmara Sea is continually injected to the heated city atmosphere by the north, northeast and southwest winds (Table 5). 

Matzarakis and Mayer (1997) calculated the values of PMV in Greece for the years 1980-1989, as in the same period of our study. But it is impossible to compare because  the results were computed by different methods in our and their studies. For the sake of comparison, the values of PMV in Istanbul for same period (1980-1989) will be examined in detail in our next study. 

Acknowledgements: This study is a part of Doctoral Thesis. “ Frequency and Intensity of Sultriness in Istanbul “ by S. Bahadır at Istanbul University. We are  thankful to Msc. Eng. E. Başak for his kindly help in preparing this article.  


Dieterichs H (1958) Dauer und Haeufigkeit schwüler stunden in San Salvador. Arch Met Geoph  Biokl B 8:369-377.

Dieterichs H (1980) Haeufigkeit und intensitaet der schwüle im hinterland der Ostfriesischen Nordseeküste.Arch Met Geoph Biokl B 28:149-164.

Fanger PO (1982) Thermal comfort. Robert E.Krieger,Florida.

Harlfinger O (1975) Vergleichende untersuchung der physiologischen waermebelastung zwischen mitteleuropa und den mittelmeerlaendern.Arch Met Geoph Biokl B 24:361-372.

Höppe PR (1993) Heat balance modeling. Experientia 49; 741-746.

Höppe PR (1999) The physiologischen equivalent temperature-a universal index for the biometeorological assessment of the thermal environment.Int J Biometeorol 43:71-75.

Jendritzky G,Menz H,Schmidt-Kessen W,Schirmer H (1990) Methodik zur raumbezogenen Bewertung der thermischen Komponente im Bioklima des Menschen.Beitr Akad Raumforsch Landesplan 114.

King E (1955) Ein empirisches schwülemass.Med Met Hefte 10:5-8.

Leistner W (1964) Die praktische bedeutung eines geeigneten schwülemasses.Arch Physik Ther,Heft 1:67-76.

Mayer H (1975) Die effektive temperatur unter dem aspekt schwüle in 2m. und 175m. über grund.Arch Met Geoph Biokl B 23:147-155.

Mayer H (1977) Schwüleverhaeltnisse an der deutschen Nordseeküste.Arch Met Geoph Biokl B 24:361-372.

Mayer H, Abele J (1977) Bioklimatische verhaeltnisse im Südwestafrikanischen   Steppen gebiet.Arch Met Geoph Biokl B 24: 337-347.

Matzarakis A,Mayer H (1997) Heat stress in Greece. Int J Biometeorol 41:34-39.

Matzarakis A,Mayer H,Iziomon MG (1999) Applications of a universal thermal index: physiological equivalent temperature.Int J Biometeorol 43:76-84.

Olaniran OJ (1982) The physiological climate of Ilorin,Nigeria.Arch Met Geoph Biokl B 31:287-299.

Table 1.  Monthly numbers of  sultry hours with S 30.0 for the years 1980-1989.


                           June   July    Aug    Sept       Total                                              

                                                                  Hours    %       

             1980         4       36       29       5         74      2.5

             1981         2       54       54       1       111      3.8

             1982       19         3       18     35         75      2.6

             1983         3      146      40       1       190      6.5

             1984       10        27      15       2         54      1.8

             1985         2        15      87     12       116      3.9

             1986     148      150    274       5       577    19.5

             1987       48      198      61     12       319    10.9

             1988       82      454    330     44       910    31.0

             1989       25      119    347     23       514    17.5

            Hours     343   1202   1255   140    2940   100.0

Total         %       11.7    40.9    42.7    4.8    100.1                    



Table 2.  Average monthly frequencies of sultry hours with S ≥ 30.0 for ten-year.


              Sultry hours / Total hours     %


JUNE           343 h /   7200 h          4.74     

JULY           1202 h /   7440 h        16.13

AUGUST       1255 h  /   7440 h        16.87

SEPTEMBER    140 h  /   7200 h          1.94

TOTAL          2940 h  / 29280 h        10.04



Table 3. Monthly total numbers of sultry hours with the classified values of sultry index S.


          Sultry index       JUNE     JULY     AUG .   SEPT.    TOTAL       % 


     S =  30.0-30.9         181      487        539       81          1288        43.8

             31.0-31.9         101      297        333       43            774        26.3

             32.0-32.9           39      202        214       11            466        15.9

             33.0-33.9           14      115          89         5            223          7.6

             34.0-34.9             3        50          46         -               99         3.4

             35.0-35.9             3        29          24         -               56         1.9

             36.0-36.9             1        15            7         -               23         0.8

             37.0-37.9             1          5            1         -                 7         0.2

             38.0-38.9             -           2            2         -                 4         0.1

                TOTAL        343     1202      1255     140          2940      100.0   



Table 4. The monthly numbers of sultry hours with S ≥ 30.0 according to the wind directions accompanied with sultriness (Vrb..=Variable). 

                             JUNE   JULY   AUG.  SEPT.   TOTAL   %


              NW            3           7         11         -           21       0.7

                  N          65       157       126         6        354      12.0

                NE          62       608       641      54       1365      46.4

                   E          20         95       145      10         270        9.2

                 SE            9         22         36        5           72        2.4

                   S          14         34         50        7         105        3.6

SW          89       154       128       23        394       13.4

                  W          36        57         56       17        166         5.6

             Calm          44        67         60       16        187         6.4

               Vrb.            1          1           2         2            6         0.2

           TOTAL       343    1202     1255     140      2940       99.9   


 Figure 1:  Hourly numbers of sultry hours with S ³ 30.0 for the years 1980-1989.

 Figure 2:  Hourly numbers of sultry hours with classified S ³ 30.0 values for the years 1980-    1989.