In Part I of this study, the development processes of Baiu frontal depressions (BFDs) have been examined through case-study numerical experiments. The numerical simulations revealed that latent heating is dominant for the development of BFDs in the western part of the Baiu frontal zone (W-BFDs), west of approximately 140°E, while both latent heating and baroclinicity are important for the development of BFDs in the eastern part of the zone (E-BFDs), east of approximately 140°E. In this study, idealized numerical simulations with zonally homogeneous basic fields are conducted to obtain a more generalized perspective of the development processes of BFDs.

 The basic fields for the idealized simulations are made from the composites of the environments under which 28 W-BFDs and 43 E-BFDs developed. The idealized simulations successfully reproduce a realistic W-BFD and E-BFD. The W-BFD has a slightly westward-tilted vertical structure, modulated by latent heating at low levels of the atmosphere. In contrast, the E-BFD has a westward-tilted structure through the troposphere, similar to the well-known baroclinic wave structure. Results of available potential energy diagnosis for the effects of latent heating and baroclinicity on the BFD development are consistent with those in Part I. The W-BFD has a mechanism mainly driven by latent heating yielding strong convection, while the E-BFD develops through baroclinic instability in moist atmosphere.

 This study investigates future changes in atmospheric circulation during the Baiu in Japan using 20-km-mesh atmospheric general circulation model (AGCM) simulations for the present-day (1979-2003) and the future (2075-2099) climates under the Representative Concentration Pathways 8.5 scenario. The simulated future climates include the outputs obtained with one control sea surface temperature (SST) and three different SST patterns. The Baiu frontal zone, defined as the meridional gradient of equivalent potential temperature, gradually moves northward during June–July–August in the present-day climate. In the future climate simulations using the control SST, the Baiu frontal zone is projected to stay to the south of Japan in June. Thus, precipitation is projected to increase over this region, while decreasing in the western part of Japan. Future changes in precipitation and atmospheric circulations in June are consistent across all four SST patterns. However, precipitation and atmospheric circulation in July and August in the future climate simulation depends on the SST patterns as follows: in non-El Niño-like SST pattern, the Baiu terminates in late July, similar to that of the present-day climate; a result with an El Niño-like SST pattern shows that sufficient amount moisture is transported to the Japanese islands and leads in a delay of the Baiu termination until August; and in the SST pattern with strong warming in the western North Pacific (WNP), a sufficient amount of moisture is transported to the south of Japan from June until August. The difference in the SST pattern leads to a variation in sea-level pressure in the WNP and affects a variation of the Northern Pacific subtropical high around the Japanese islands in July and August.

 Mechanisms related to the diurnal cycle of tropical deep convection over a complex terrain were investigated in the Bandung basin, West Java, Indonesia. Observational data were analyzed from X-band radar, Global Navigation Satellite System (GNSS) receivers, and radiosondes, in conjunction with high-resolution numerical model data.

 Significant diurnal variation of GNSS-derived precipitable water vapor (PWV), which peaked in the early evening, was observed from 13 to 19 March 2013. During this period, the X-band radar detected convective initiation at approximately 1200 local time over the southern slope of the basin. A 2-km-mesh model successfully simulated the observed diurnal variations of PWV and rainfall from 15 to 17 March 2013. In the model, moist air was present at the bottom of the basin early in the morning, which was transported to the southern slope of the basin by valley wind circulation after sunrise. In contrast, humidity was lower in the northern part of the basin due to a downward circulating valley wind. The valley wind decreased static stability around the southern slope of the basin by transporting moisture. It also caused a low-level wind convergence, resulting in convective initiation on the southern slope of the basin. The GNSS receiver network also recorded this simulated water vapor variability associated with the valley wind.

 These results suggest that water vapor in the bottom of the basin during mornings and its advection by the valley wind strongly influences convective initiation in Bandung.