Seasonal variations in the cooling effect of Barcelona's urban parks: an approach using remote sensing

Abstract

Urban parks are spaces of opportunity for the adaptation of cities to climate change. Its commonly abundant vegetation and unsealed surfaces, break the continuity of the artificial land cover that distinguishes urban areas, and generate a temperature reduction in relation to its context, which extends towards its surroundings and generates the so-called urban park cool island (PCI). Therefore, recognizing the role of the physical characteristics that define this effect is an opportunity to propose parameters of climate adaptation of cities and reduce health risks during extreme heat wave episodes. In recent years, the addition of the Land Surface Temperature (LST) to studies of the climatic behavior of urban spaces, has allowed addressing a greater number of cases with simultaneous thermal information in larger-scale areas. In this context, in this work we propose a remote sensing approach to the quantification of the seasonal variations of the cooling effect of the urban parks of Barcelona and the role of their design and location. Particularly, we highlight the relationship of the annual changes in the composition of the green spaces inside and around the parks with their microclimatic behavior. For this purpose, we calculate the PCI limit extension (Lmax) and intensity (ΔTmax) of 86 urban parks in the city of Barcelona, through a method of homogenization of the surrounding temperature using concentric rings, based on the LST of Landsat-8 OLI / TIRS satellite of representative days of the four seasons of the year. Subsequently, we analyze the correlation between these indicators and the physical attributes of shape, location and annual variations in the composition of the green spaces of the parks and their surroundings. Where we obtained the latter by quantifying changes in the Normalized Difference Vegetation Index (NDVI) from the same satellite images.

First results record that the urban parks in Barcelona has an average LST per season between 0.99 and 2.56ºC lower than the urban context. Where we register average PCI values of 57.91 to 78.02m of Lmax and 0.75 to 1.70°C of ΔTmax. In addition, throughout the year, 83 parks register positive PCI and only one negative the entire year. Most of the parks record their maximum PCI during summer and their minimum in winter. On the other hand, the set of parks records an average NDVI of 0.17 to 0.22 greater than the urban context of Barcelona. Where, the annual variation range of NDVI of each park does not correlate with its PCI variations. In addition, coinciding with the literature, ΔTmax is the indicator of the PCI with the highest correlation with physical attributes throughout the year. Particularly, the greater the proportion of areas with clustered trees (identified with an NDVI from 0.50 to 0.60), the greater the ΔTmax in the four seasons, while the rest of the attributes vary in each season. As for the Lmax, it does not register a constant significant correlation during the year with some physical attribute, although with the ΔTmax does. In general, the results point to a predominant influence of the physical characteristics of the park shape and the composition of its green spaces. Leaving aside its location and the characteristics of its urban context.

Given this, as conclusions we discuss the relevance of the methodological approach to the case studied and the contributions to the recognition of the influence of the design and location of urban parks in Barcelona on their microclimatic influence. The present study is part of the project “Urban-CLIMPLAN. The urban heat island: effects on climate change and modeling for territorial and urban planning strategies. Application to the Metropolitan Region of Barcelona”; financed by the Ministry of Economy and Competitiveness of Spain (MINECO) and the European Fund for Regional Development (FEDR).