Assessment of Convection-Permitting Simulations in Representing the Urban-Rural Contrasts over Selected Cities in South America

Abstract

This study focuses on assessing the capability of simulating the urban-rural contrast in three cities: Buenos Aires, Córdoba and Porto Alegre. Coordinated 3-year long simulations covering subtropical South America performed with six convection-permitting regional climate models (CPRCMs) at 4 km resolution, two simulations performed at 12 km resolution and one long-term CPRCM simulation covering the CORDEX South American domain were used. The set of models is characterized by different complexities in the urban canopy treatment; therefore, the impacts of resolution and model complexity are evaluated. All CPRCMs capture warmer, drier and weaker wind conditions over the cities and more intense heat waves, ranging from 0.4 °C to 2.1 °C for Buenos Aires, 0.3 °C to 1.5 °C for Porto Alegre and up to 1.9 °C for Córdoba, being the observed signal within the simulated range. The 12 km runs capture the main urban imprint, albeit with reduced magnitude and spatial detail compared with CPRCMs. The definition of the urban mask emerged as one of the key parameters modulating the characteristics of the simulated urban signal, particularly for smaller cities. For Buenos Aires, with a similar representation of the urban and rural areas from the set of models, complexity on the treatment of the urban canopy improve the simulated intensity of the city signal and capture a non-uniform distribution of the urban signal within the city. These results guide the selection of appropriate model configurations for urban climate studies and inform future efforts to assess climate risks and adaptation needs in rapidly growing South American cities.