A half-automated method was designed to identify satellite signals over areas where there is persistent cloud growth that leads to isolated storm initiation. Successive (30-min interval) and gridded (0.05° × 0.05°) infrared (~10.8-) images were utilized to produce candidate satellite signals involving grid(s) with signatures of freeze transition and a persistent cooling rate of no less than 8°C per 30 min. All candidate satellite signals were then validated with radar mosaics to identify if they actually represent relevant growing cumulus clouds undergoing isolated CI.
A total of 1630 satellite signals were identified to have corresponding isolated-CI processes with an increasing monthly mean frequency from 2 per day in May to 39 per day in August. Satellite signal formation predominantly peaks in the early afternoon when the solar heating is strongest. The majority of identified signals prefer to occur over the areas with remarkable inhomogeneity over CEC such as mountain areas, water areas, and areas of mountain water combined. Due to the northward movement of monsoon rainband and the moister environment with increasing conditional instability as the summer progresses, there is a shift of signal-preferred areas from northwest CEC (mainly uplands) in the early summer to southeast CEC (mainly lowlands) in the midsummer. All these results are reconfirmed when performing interannual examination, highlighting the impacts of local, boundary layer processes on isolated storm initiation.
The satellite signals have longer lead time for isolated CI in late morning and afternoon than in early morning and nighttime. During the nighttime and early morning, the distinction between cloud-top signatures and background terrestrial radiation became less apparent, resulting in delayed identification of satellite signals and thus short lead time. The lead time shows a declining trend from May to August. As the summer progresses, the cumulus clouds generally develop in a moister environment with increasing background instability, and thus require less time to develop into the altitude enabling precipitation onset due to the stronger cloud development and increasing warm-rain process. In CEC, the lead time overall increases with the landscape elevation of isolated CI. This is likely due to the dominant warm-rain process over sea and plain, along with slower cloud growth from hill and mountain to plateau. It is envisioned that these results would have important implications for CI nowcasting in China and contribute to the CI knowledge in similar climate regimes around the globe.