Gap-filling method of measured wind data for passive cooling in Indonesia

Typical Meteorological Year (TMY) data, constructed for 106 locations in Indonesia in 2024, lacks wind direction components, which are essential for passive cooling design in buildings and promoting energy efficiency. This study addresses gaps in wind data using bias-corrected ERA5 reanalysis data and the Monte Carlo method. Wind data from 106 Indonesian stations (2011-2020) had 30-48% missing hourly values, which were addressed through three proposed gap-filling techniques: (1) Using ERA5 wind speed-debias and the Monte Carlo Method to simulate wind directions; (2) Retaining original ERA5 U and V components, applying bias corrections, and converting back to wind speed and direction; and (3) Applying the same technique to ERA5-Land data, with higher spatial resolution. Model verification involved wind rose diagrams, circular correlation, RMSE, and MBE. The first technique showed the best performance in 106 locations with a correlation range of 0.09 to 0.8, an average RMSE of 64°, and an average MBE of 40°. This technique was chosen to fill gaps in wind direction data, reconstructing TMY wind direction data. Results indicated dominant wind directions from East (19.8%), West (15.09%), South (12.26%), and North (11.3%) with significant regional variations in Indonesia.