The estimation of sea-breeze front velocity over coastal urban using Himawari-8 images: A case study in Jakarta

Abstract

The sea breeze is a meteorological phenomenon that occurs due to the contrast temperature between land and oceans. The propagation velocity of sea breeze are influenced strongly by e.g., synoptic wind and geographical conditions. Therefore, it is important to understand the relationship between the spatial distribution of sea breeze velocity and the surface characteristic, for instance over urbanized and less-urbanized coastal areas. When the sea breeze propagates inland, a cumulus cloudline will form in the vicinity of the sea breeze front (SBF). Previous studies have successfully detected the cloudline automatically using the morphological-snake algorithm. In this paper, we estimate the SBF velocity using Himawari-8 satellite images. The proposed method segmented the cloudline data points using a clustering approach, named machine learning-based k-means++, on the level-set obtained from snake algorithm. We then estimate the SBF velocity by calculating the haversine distance of the segmented cloudline points that propagate over time. The comparison of estimated cloudline speed with SBF speed measured at two observation sites, namely KKP and BPL, reveals the root mean square errors 1.39 m/s and 1.41 m/s, respectively. And the propagation direction was mainly southward.