PLEA2007 - The 24th Conference on Passive and Low Energy Architecture, Singapore, 22-24 November 2007 The Impact of Urban Vegetation on Microclimate in Hot Humid São Paulo Jörg Spangenberg 1,2 , Paula Shinzato 1 , Erik Johansson 1,3 and Denise Duarte 1 1 Laboratory of Environment and Energy Studies (LABAUT), University of São Paulo, Brazil 2 Bauhaus University, Weimar, Germany, 3 Housing Development & Management, Lund University, Sweden joerg_spangenberg@yahoo.com.br, paulashinzato@yahoo.com, erik.johansson@hdm.lth.se dhduarte@terra.com.br ABSTRACT: Field monitoring in a park, a square and a street canyon on a summer day in São Paulo, Brazil, showed that the park was up to 2K cooler than the square and the canyon. The effect of adding shading trees to the street canyon was simulated for the same day using the numerical model ENVI-met. The simulations showed that incorporating street trees in the urban canyon had a limited cooling effect on the air temperature (up to 1.1K), but led to a significant cooling of the street surface (up to 12K) as well as a great reduction of the mean radiant temperature at pedestrian height (up to 24K). Although the trees lowered the wind speed, the heat stress was mitigated considerably as the physiologically equivalent temperature (PET) was reduced by up to 12K, mainly due to shading. Keywords: urban vegetation, urban microclimate, numerical simulation, pedestrian comfort 1. INTRODUCTION Vegetation is an important design element in improving urban microclimate and outdoor thermal comfort in urban spaces in hot climates [1]. Due to urbanization, however, vegetation is scarce in many tropical cities. The main benefits of vegetation in hot climates are reduced solar radiation and lower air temperature due to shading and evapotranspiration [2]. Lower air temperatures are essential both to improve thermal comfort conditions of pedestrians and to limit energy use for cooling. This paper shows results from an ongoing study of the benefits of vegetation in the city of São Paulo, Brazil. This 19 million sprawling metropolis is characterised by a heterogeneous urban structure as well as by an uneven distribution of urban green. It is located at 23°32’S, 46°37’W, 60km from the sea at 720-850m altitude. The city’s summer climate is hot and humid with mean temperatures between 22 o C and 30 o C, but temperatures above 30 o C are common. 2. METHODOLOGY Climate monitoring was carried out on 19 th of December 2006, a typical summer day, in the central area of Luz. An 8 ha park, a square and an urban canyon were compared in order to examine the effect of the park’s vegetation. Three meteorological stations (Huger WM 918 and 968, as well as ELE MM900) were used to measure air temperature, humidity and wind speed simultaneously at 1.10m height at the three locations. The urban canyon, which is located in a dense area with an average height-to-width ratio of 2.2 and almost devoid of vegetation, was chosen for a parametric study using the microclimate model ENVI- met [3]. The ENVI-met model, which was chosen due to its advanced approach on plant-atmosphere interactions, was carefully calibrated with the meas- urement results by adjusting various input parameters related to local climate and urban surface properties so that similar daily temperature and humidity curves were achieved. The comparison between measured and simulated air temperature is shown in Figure 1. In addition to the actual situation, two different greening scenarios were simulated, one with less dense street trees (T1) and another with a much denser tree canopy (T2). The leaf area indices (LAI) for T1 and T2 were 1 m²/m² and 5 m²/m² respectively Both tree types were 10m high with canopies covering the entire street. Both parametric tree models had ellipsoid leaf area distributions with maximum Leaf Area Densities (LAD) located in the middle of the crowns. The attenuation of solar radiation at midday was 39,3% for T1 and 91.9% for T2. The aim was to evaluate how the LAD of the two tree canopy types influenced air temperature, surface temperature, humidity, wind speed and, consequently, thermal comfort. The Physiologically Equivalent Temperature (PET) was applied to this study. This comfort index was calibrated by Monteiro and Alucci [4] for open spaces in the city of São Paulo interviewing approx- imately 2.000 people adapted to local conditions. 3. RESULTS The simulated results are shown for the crossing between the east-west and north-south oriented street canyons.