VARIATIONS AND TRENDS OF THERMAL COMFORT AT THE ADRIATIC COAST

Transcription

1 VARIATIONS AND TRENDS OF THERMAL COMFORT 7 AT THE ADRIATIC COAST Ksenija Zaninovic 1 and Andreas Matzarakis 1. Meteorological and hydrological service, Climatological research department, 1 Zagreb, Gric 3, Croatia. Meteorological Institte, University of Freibrg, Werderring 1, D-7985 Freibrg, Germany address: zaninovic@cirs.dhz.hr (Ksenija Zaninovic) ABSTRACT Knowledge abot the thermal comfort of hmans plays an important role in health and activities, especially in torism, recreation, leisre and sport. In spite of the efforts that have been made to investigate temperatre changes dring the last centry, reslts cannot completely clarify their impacts on hmans. This paper analysed variations and trends of the Physiologically Eqivalent Temperatre (PET) and Predicted Mean Vote (PMV), two thermal bioclimate indices based on hman energy balance models. Climatic changes were analysed sing data from the period of 187-, taken from the meteorological station at Hvar,Croatia, a poplar torist destination island in the Adriatic Sea. This analysis of thermal bioclimate conditions has showd that changes in the examined period were within the range of one class of physiological strain for hmans. KEYWORDS: Thermal bioclimate, Binomial moving average filter, Trend analysis, Progressive analysis INTRODUCTION Until now climate change in Croatia has been investigated sing minimm and maximm temperatres, daily temperatre range and precipitation from inland lowland and coastal stations. The reslts indicate decreasing maximm temperatres and increasing minimm temperatres, leading to a significant decrease in daily temperatre range (1). Hvar was not inclded in those investigations de to missing data for certain years. However, the changes in thermal comfort for Hvar (Fig. 1) have been investigated for the periods of (), and 191- (3). The reslts for the first show increasing thermal comfort trends being significant in winter, atmn and annally as the reslt of positive temperatre trends and negative wind speed trends. The analysis

2 75 for the second period show a negative trend for thermal indices in all seasons as a reslt of the increasing trend in wind speed, despite positive temperatre trends. Similar investigations have been made for the montainos meteorological station Zavizan (159 m), for the shorter period of (). The trends of thermal comfort were positive in all seasons as well, bt significant for smmer, atmn and annally. HVAR Figre 1: The location of Hvar METHODS.1 Thermal environment The thermal effective complex deals with the inflences of the thermal environment on the wellbeing and health of hman beings. The basis for this is the close relationship between the hman thermoreglatory mechanism and the hman circlatory system. For the physiologically significant assessment of the thermal environment, some thermal indices are available which are derived from the hman energy balance (5,, 7, 8, 9). Several investigations have been performed which se thermal indices, sch as PMV or PET, for the hman-biometeorological assessment of the thermal environment in different scales. Reslts from case stdies (9) enable a process analysis, e.g. in the form of regressions between PET and meteorological inpt parameters sch as single radiative flxes, mean radiant temperatre, air temperatre, vapor pressre and wind speed. For calclating the mean radiant temperatre, the hman-biometeorological radiation model RayMan (1) was sed, which is well sited for application in applied climatological and meteorological stdies.

3 7.. Trend analysis The flctations and trends of seasonal and annal vales of the thermal comfort indices PET and PMV, as well as the meteorological parameters that inflence thermal comfort (air temperatre, relative hmidity, wind speed and clodiness), were determined. Variations and trends were analysed dring the period of 187-, in spite of some missing data in the s and s. In order to remove short-term flctations the data series was smoothed by means of the weighted 11- year binomial moving average filter. The linear trend has been tested for significance by means of the nonparametric Mann-Kendall rank statistics t (11, 1). For the series, which showed the significant trend identified by the Mann- Kendall coefficient t, a progressive analysis of the time series by means of the statistic (t) was performed in order to determine the beginning of this phenomenon by means of a seqential analysis (1). RESULTS According to the mean annal and seasonal PET and PMV vales, the mean annal thermal sensation in Hvar from 187- was slightly cool (17. C PET, -.8 PMV), varying from cool winters (.5 C PET, -3.1 PMV) to slightly warm smmers (3.7 C PET, 1. PMV). Becase of the maritime inflence, atmn was warmer (19.3 C PET, -. PMV) than spring (15.9 C PET, -1. PMV). The PET and PMV flctations showed a visible warming at both the beginning of the centry and arond the 195 s (Fig. ), related to decreases in wind speeds in the same periods. After the warming in the 5 s, a cooling period occrred ntil the beginning of 8 s, as the reslt of a decrease in temperatre and simltaneos increase in wind speed. The warming in PET and PMV from the beginning of the 8 s ntil the end of centry was the reslt of an increase in air temperatre and decrease in wind speed, bt also a decrease in clodiness from the end of the 7 s ntil the end of the 8 s. Both hman-biometeorological indices, PET and PMV, showed increasing trends in all seasons, significant for winter, atmn and annal vales. These positive trends were the reslt of increasing temperatre and decreasing wind speed. The greatest change was temperatre in winter (arond. C/1 years), and the smallest was temperatre in spring (. C/1 years). However, only the increasing trend of mean annal vales of. C per 1 years was significant (Tab. 1). The wind speed decreasing trends were significant for the winter, atmn and annal vales, the same as for the hman biometeorological indices. Vapor pressre also contribted to the increasing trend in thermal sensation, becase of increasing trends in all seasons (althogh statistically insignificant).

4 Finally, clodiness showed positive trends in all seasons, and only the winter trend was not significant. 77 C Physiological eqivalent temperatre 18 1 PET = (t-187) C Air temperatre 18 1 t = (t-187) m/s Wind speed v = (t-187) hpa 1 Vapor pressre 1 1 VP = (t-187) Tenths Clodiness 5 C = (t-187) Figre : Annal variations of the physiologically eqivalent temperatre ( C), the air temperatre ( C), wind speed (m/s), vapor pressre (hpa), wind speed (m/s) and clodiness C (in tenths), inclding a weighted 11-year binomial moving average series, and linear trends dring the period of 187- at Hvar

5 78 Table 1: Seasonal and annal trends (per 1 years) of mean physiologically eqivalent temperatre (PET in C) and predicted mean vote (PMV), temperatre (t in C), vapor pressre (VP in hpa), wind speed (v in m/s) and clodiness C (in tenths). Shading denotes trends significant at the.5 level according to Mann-Kendall rank statistics. Period: 187- Seasons PET PMV t VP V C Winter,7,1,3,8 -,1,59 Spring,35,8,17,1 -,39,81 Smmer,5,1,3, -,11,73 Atmn,99,,3,1 -,5,39 Annal,8,1,35, -,,58 The progressive trend test was applied to the annal vales of PET, which had significant increasing trend, as did the parameters inflencing thermal comfort - temperatre, wind speed and water vapor pressre (Fig. 3). From the graphical representation of the onward () and backward ( ) test series of PET it can be seen that dring the analysed period they overlap several times, implying the absence of a trend. The last intersection point between and occrred in 1988, while exceeded the 1.9 limit vale in 199, sggesting the beginning of a significant positive trend. However, as the effect was very recent it is advisable to await confirmation from ftre observations, especially becase of the many changes in the previos periods. The progressive trend test for temperatre shows that the increasing trend in temperatre began in 19. The increasing trend of temperatre became significant in 1959 and, besides some flctations in the 7 s, stayed significant ntil the end of the observing period. In spite of some similarities between PET and temperatre trends, it is obvios that variations in the PET trend were the reslt of other meteorological parameters important for thermal comfort. The progressive trend test for wind speed showed that, in spite of a significant negative trend for the whole period, wind speed changed from the beginning of the 3 s. However, the intersection between onward and backward test series for wind speed in 197 cannot be taken as the beginning of an increasing trend, bt it is obviosly corresponded with the retard in the increase of PET. Finally, the increase of PET in the last decade of the centry was the reslt of simltaneos sharp increases in temperatre.

6 79 Physiological eqivalent temperatre ' Air temperatre 1-1 ' Wind speed ' Vapor pressre ' Clodiness ' Figre 3: Progressive trend test for annal vales of mean physiologically eqivalent temperatre, air temperatre, wind speed, vapor pressre and clodiness at Hvar, dring the period of 187-

7 8 CONCLUSIONS For the assessment of the thermal environment on hman beings in different scales, hmanbiometeorology provides well-sited thermal indices on the basis of the hman energy balance. Investigations of thermal bioclimate for the qantification of the effects of atmospheric conditions on hman beings reqire long data series to check trends, and to see if the trends are significant. The analysis of climate change throgh thermal indices in Hvar from the middle 19 th centry showed a positive trend in all seasons as a reslt of positive temperatre trends and decreasing trends in wind speed. This reslt coincides with earlier investigations into the trends in thermal comfort at Hvar, calclated with a different thermal comfort index (). On the other hand, the analysis of thermal comfort changes in Hvar from the beginning of th centry showed the opposite trend as the reslt of increasing wind speed trend in the centry (3). Trends of the bioclimatic conditions of torism areas provide information for the torism indstry and govermental athorities, allowing adeqate planning for the expected changes in the natre and length of the torism season. REFERENCES 1. Zaninovic, K. and Gajic-Capka, M. Extreme Temperatre Changes in this Centry in Croatia. Hrv. Meteor. Cas. 3:1-.. Zaninovic, K Seclar Variations and Trends of Thermal Comfort at the Adriatic Coast. Book of abstracts IGU, Climate and Environmental Change. 1998, -3 Agst, Evora, Portgal, Zaninovic, K. and Matzarakis A.. Climatic changes in thermal comfort at the Adriatic coast, Advanced Research workshop Climate change and Torism. Warsaw (in print). Zaninovic, K Trends of Thermal Comfort on Dinaric Alps. Proceedings of the th International Conference on Alpine Meteorology ICAM 9. September 9-13, Bled, Slovenia, 199: Mayer, H Urban bioclimatology. Experientia. 9: Matzarakis, A. and Mayer, H Heat stress in Greece. Int. J. Biometeorol. 1: VDI, Methods for the hman-biometerological assessment of climate and air hygiene for rban and regional planning. Part I: Climate, VDI gideline Part. Beth, Berlin, Höppe, P The physiological eqivalent temperatre a niversal index for the biometeorological assessment of the thermal environment. Int J Biometeorol. 3:71 75.

8 81 9. Matzarakis, A., Mayer, H. and Iziomon, M.G Applications of a niversal thermal index: physiological eqivalent temperatre. Int. J. Biometeorol. 3: Matzarakis, A., Rtz, F. and Mayer, H.. Estimation and calclation of the mean radiant temperatre within rban strctres. WCASP-5, WMO/TD No. 1: Mitchell, J. M.Jr. et al. 19. Climatic Change. WMO. Tech. Note No Sneyers, R On the Statistical Analysis of Series of Observations, WMO, Technical Note No 13.