An observational study of variation in convective features in Palau: transition from diurnal to 2-day periods

We present a case study of diurnal to quasi-2-day variability in convective features over the open ocean with environmental modulation in a convectively suppressed period. A 20-day period of high-frequency radar observation data was analyzed using radiosonde and flux data around the Republic of Palau in the western Pacific warm pool. The convective features were defined as continuous radar echoes and were classified into three horizontal size categories. The number, area, and rainfall of the small and medium size categories of convective features exhibited significant diurnal cycle signals, with maxima occurring from midnight to predawn during the suppressed period. This time of maximum rainfall was consistent with most previous studies, but was inconsistent with the afternoon maximum rainfall of the Tropical Ocean and Global Atmosphere Coupled Ocean-Atmosphere Response Experiment (TOGA COARE). Different mixed layer evolution was responsible for the different timing of maximum rainfall in these two experiments. Variability in the number and area of convective features was studied intensively during a 5-day period lasting from the suppressed period to the active period. Variability was modulated from diurnal cycles to 2-day cycles. The convective features were always initiated in the evening by mixed layer evolution; however, the timing of dissipation was delayed in the later cycles. Moisture accumulation in the lower troposphere appeared to modulate the duration of convective evolution in later cycles. This modulation from diurnal to 2-day cycles was understood as a transition from diurnal cycles in the suppressed period to quasi-2-day cycles in the active period found in TOGA COARE. Lower tropospheric moisture had a clear correlation with the number of small- and medium-sized convective features. These convective features moisten the lower troposphere, regulating the evolution of subsequent organized convection.