The micro-climate of a mixed urban parkland environment

Progression of climate change, with its predicted intensification of temperature extremes and heat wave durations, combined with demographic trends towards increased urbanization makes the study of urban micro-climates desirable. Understanding of mixed urban parkland morphologies leads to insights into possible adaptation and mitigation strategies to minimize impacts due to temperature extremes and UHI (urban heat island) on human health. Observational methodologies to study these environments present difficulties in obtaining data of sufficient spatial and temporal resolution, and are expensive and time consuming as well. Modelling using mathematical computer simulations addresses some of these concerns. However, confidence in the results obtained from models requires verification of accuracy. Merely observing that the modelling output looks plausible isn't enough. Verification of underlying processes and their interactions are ultimately necessary for complete confidence. Data collected in a mixed urban parkland study area was analysed for spatial and temporal temperature variations. Urban micro-climate drivers such as incoming shortwave radiation, wind, and humidity played a role in the variations across the area. Wind was found to be an important driver. It moderated afternoon maximums through mechanical mixing at solar exposed sites as effectively as tree cover shading did at other sites. At the same time, heat was allowed to build at wind sheltered sites. Calming winds also contributed to dropping temperatures after dusk and warming temperatures in pre-dawn hours coinciding with increasing wind speeds. On average, temperatures in parkland areas were found to be 2°C cooler than urban areas. Modelling of this study area was carried out using ENVI-met, a urban micro-climate model. However, ENVI-met's ability to predict the temperature gradients seen in the observations was hampered by constant values, both spatial and temporal, in wind speed and humidity levels. As these were found to be important components in driving spatial temperature variability in the observations, these constant values are unable to drive variabilities as they would have in the observations. Temperature variations lag behind observed values and temperatures are also underpredicted during the day and over-predicted at night. This leads to low confidence levels about ENVI-met's accuracy in resolving temporal and spatial temperature variation and yields an inconclusive result as far as modelling predictions are concerned.