Abstract
Abstract CLIMATIC STUDY OF A GREEN SPACE OF LISBON: THE CALOUSTE GULBENKIAN FOUNDATION PARK Measurements of various climatic parameters were carried out in Calouste Gulbenkian Foundation Park, an average-sized green space (8.5 ha) in central Lisbon (fig. 1). The objectives were twofold: a) to measure the thermal differentiation between the park and the neighbouring urban area (local scale) during the day and at night; and b) to analyse the daytime microclimatic differentiation within the green space itself. The first part of the paper is a synthesis of the main influences of the green areas in urban climate; green spaces contribute to attenuating the effects of the Urban Heat Island, which is particularly important in a context of global warming. The climatic specificities of green spaces are mainly due to the reduction of incident solar radiation at the ground level under tree covering and to the increment in latent heat flux due to evapotranspiration. Contrasts within green areas create a mosaic of microclimatic conditions which are favourable to a variety of uses under different weather conditions. In order to achieve the former objective (local scale), several air temperature measurement devices were set up both within and outside the park (fig. 2), at a height of around 3m. The air temperature was recorded every 10 minutes. The analysis of the results was made using the average value for each 30 minute period. The sensors gathered data over 117 days in the summers of 2004 and 2005, as well as during 10 days in the autumn and 12 days in winter. In order to analyse the differences between the park and the surrounding built-up area, the park cool island was calculated: Dif maxt = Tmaxut – Tminvt, with Tmaxut and Tminvt as the maximum and minimum temperatures in the built-up area and the park, respectively; besides estimating Dif maxt, the median of the differences between the temperatures in the built-up area and the park at any given moment, which was designated as the median park cool island (Dif medt), was also calculated. At the microclimatic scale, six itinerant measurements of the air temperature, relative humidity, wind speed and solar and infra-red radiation were carried out in several different micro-environments within the park. All the measurements were made during daytime, the period in which the park is subject to greatest use. In order to integrate all the atmospheric parameters in a physiologically significant way, the Physiological Equivalent Temperature (PET) was calculated. The local scale study confirmed that the Park is cooler than the surrounding built-up area both during the daytime and at night and in all the seasons, but especially so in the summer and autumn (tables V and VI); in winter time, the differences were usually much less significant. It should be borne in mind that the measurement periods were much shorter in the autumn and winter than in the summer. Therefore, the results for the latter seasonal period are much more consistent. Moreover, due to the fact that the particular period in the autumn in which the measurements were carried out was very hot and dry, the thermal behaviour was in fact very similar to that of a summer period. The daytime differences were more significant than the nighttime ones. In the summer daytime, the median differences were most commonly between 1 and 1.5ºC, although extreme values of Dif maxt above 9ºC could also be found; these extreme differences occurred under very hot and dry weather conditions. The largest nighttime differences also occurred under hot and (mainly) dry conditions. This demonstrated the importance of the urban green spaces in terms of locally mitigating the heat waves, which can have dangerous effects upon the human health. The aforementioned daytime differences between the park and the surrounding built-up area were computed without taking into account the exposure to direct solar radiation. Obviously, the places in the sun were always hotter than those in the shade, regardless of being located inside the park or not. In order to find out the extent to which the park contributes to the cooling of the atmosphere, independently of the solar exposure effect, a comparison was made between measurement places under identical conditions of exposure to direct solar radiation inside and outside the park (figures 3 and 4). The sites inside the park were always cooler than the sites outside: under the sun (fig. 4), the temperature inside the garden was 2 to 3º C lower than that in the built area; the differences between the places in the shade were even larger, reaching 5ºC. During the nighttime (fig. 8), the level of thermal differentiation was found to be lower, but still the park was almost always cooler than the built-up area in all the seasons. As has been amply discussed, the night time urban temperatures are highly dependent upon the Sky View Factor (SVF). This parameter is often higher in green areas, although this is not a general rule. It is important to understand the extent to which the lower temperatures that can be found inside green areas are merely a consequence of a higher SVF (in which case it is the geometry of the place that is important, rather than effect of being located inside a green space). By comparing two measuring sites, J2 (inside the park, with dense arboreal coverage and SVF = 0.16) and E4 (in an open space outside the park, with SVF = 0.75), it was found that the former site (J2, inside the park) was almost always cooler than E4, with a statistically significant median difference of 0.8ºC (F=27.3 to df = 1, 2220; probability level = 0.95). This indicates that the effect of being located inside the green space is more important for the thermal environment than the SVF. At the microclimatic scale, the daytime differences between several measurement sites inside the park were analysed. As might be expected, the most significant differences were found to occur with regard to solar radiation (and, consequently, mean radiant temperature), while microclimatic variations of infra-red radiation, wind speed and air temperature, though present as well, have a less significant impact upon the bioclimatic conditions. The solar radiation flux in the shaded areas averaged 12% of that recorded in the places subject to direct sunlight. The differences in the air temperature were much smaller (only around 2ºC), but the thermophysiological result clearly demonstrates the importance of shade in the summer days (the mean difference in the PET was 18.5ºC). Moreover, it was found that groups of trees cause a much greater reduction in solar radiation than isolated trees: in one site under a group of trees, a reduction in solar radiation of 94 % was recorded, and the mean Tmrt was 26ºC; in the same period, under an isolated tree, the reduction in solar radiation was only 76 % and the mean Tmrt was 33ºC. This difference was mainly due to the lateral fluxes of diffuse radiation, which under isolated trees were 9 times as high as those under groups of trees, whereas the vertical fluxes (which penetrate through the tree top) were only 4.1 times as high.