Benchmarking urban local weather with long-term monitoring compared with weather datasets from climate station and EnergyPlus weather (EPW) data

Abstract Weather is essential for building energy simulation and energy efficient retrofitting since building energy use is highly weather-dependent. Weather data provided by EnergyPlus weather (EPW) data or generated by urban climate stations are used in most simulation tools. However, weather varies temporally and spatially, and the gaps may exist in the typical-year weather, weather from peri-urban areas, and local actual weather. This paper compares three types of weather datasets, from EPW data, from an urban climate station (STATION), and from long-term measurement with a local micro-climate station at Southeast University (SEU) campus. Results show there are significant gaps between the three weather datasets and compared with the STATION and EPW data, the average temperature in the micro-climate (SEU) is higher by 1.2 °C and 2.2 °C, respectively. The cooling degree day in SEU weather data is 20.4% and 40.8% higher, with heating degree day 15.7% and 26.3% lower. Comparing average relative humidity shows a big difference of 18.3% and 22.1% between SEU, STATION and EPW, and the former is comparatively lower than the latter two and no clear correlated trend. The SEU weather data has 23.1% more solar radiation hours than STATION weather data. As for wind speed, the average wind speed in the SEU weather data is lower by 1.14m/s and 1.29m/s, respectively, when compared with the STATION and EPW data. This study could be the reference for different trials for energy performance dynamic simulation for building energy analysis.