Regional Variability of Raindrop Size Distributions in Indonesia as Inferred from Principal Component Analysis

Distribution of precipitation in Indonesia is very unique due to a complex interaction between local factors such as complex shape, geography, and orientation of the island coastlines and global scale atmospheric circulations. Raindrop size distribution (DSD) which has important implications for remote sensing and several numerical modeling applications may also vary considerably across the region. This work provides new insights into precipitation microphysics in Indonesia as inferred from principal component analysis (PCA) with 7 attributes. The PCA is used to classify raindrop size distribution data collected through a network of the Particle Size and Velocity (Parsivel) optical disdrometer along equatorial Indonesia. The Par-sivel were installed at Manado (Sulawesi) and Biak (Papua). In addition to the aforementioned sites, we have been conducting DSD observations at Padang, Sicincin, Kototabang (Sumatra) and Pontianak (Kalimantan). We divided the data into 6 DSD groups indicating six unique DSD characteristics. It was found the percentage of DSD in group V (large Do and low N
_w
) at Kototabang, Sicincin, Padang and Pontianak is larger than that at Manado and Biak. On the other hand, the percentage of data point in group IV (small Do and large N
_w
) and VI (small Do and very large Nw) at Manado and Biak is smaller than other locations.