Reference weather data selection in urban weather generator model

Abstract

The precision and degree of detail of input urban area parameters and meteorological files directly affect the final result of a simulation. The objective of the research is to evaluate the performance of the Urban Weather Generator v4.1 model (UWG) for urban climate analysis in two cities. Experimental measurements during the summer period in two cities (Mendoza -desert climate- and Campinas -warm temperate-) were used to verify the accuracy of the model. A comparison with two types of weather files (.epw) was used to verify the performance of the input data. The Typical Meteorological Year (TMY) weather data, which is the weather file commonly used with this software, and a weather data based on the year of collection of the experiment (Single_Year) were used.This research showed how the use of the one-year measurement weather file (Single_Year) greatly reduces errors between observed and simulated data. The air temperature registers an R2=0.96 (Mendoza) and 0.98 (Campinas). Whereas if the simulation is performed with typical long-term weather data (TMY), the coefficient of determination is null (R2 = 0.43 in Mendoza and R2 = 0.11 in Campinas). The same was verified with the relative humidity in both cities. The underestimation or overestimation of temperature and humidity leads to errors in calculating the use of equipment for interior conditioning of buildings. Therefore, to obtain accurate results at a given location, it is necessary to know the local climate at short intervals, usually at an hourly rate, during a given year.