Diurnal Cycles of Tropical Convective Processes in Satellite-Observed, Reanalysed and Simulated Frozen Water Paths

In Earth System Model (ESM) simulations, estimates of Frozen Water Path (FWP), i.e., the column-integrated mass of precipitating and suspended ice particles, exhibit large uncertainties. In tropical regions with prevailing deep convection, FWPs are further characterised by a strong diurnal variability in simulations and observations. However, evaluating the simulated diurnal cycles has been difficult due to a lack of long-term satellite observations. Here, we use the novel machine-learned Chalmers Cloud Ice Climatology (CCIC), based on merged satellite datasets, to explore potential deviations of the captured diurnal cycle of FWP, both in ERA5 and km-scale models of the DYAMOND project. Moreover, we crosslink the diurnal cycle of FWP with the ones of precipitation and high-cloud cover to gain a broader view of the diurnal cycle of deep convection.  We find a general agreement on the phase of the diurnal cycle of FWP in CCIC and DYAMOND km-scale models. In contrast, ERA5 shows shifted FWP diurnal cycles over all evaluated tropical regions. Both DYAMOND models and ERA5 underestimate the diurnal amplitude of FWP and overestimate the diurnal amplitude of precipitation. Diurnal cycles of the observed variables are characterised by pronounced land-ocean contrasts. Tropical land areas show a year-round afternoon peak of precipitation, which is closely followed by a peak of FWP, while high cloud cover peaks are delayed towards evening or midnight, depending on the season. Tropical oceans have a broad peak in high cloud cover in the evening hours. This is followed by a build-up of both precipitation and FWP over the night towards an early morning peak. These findings indicate that different processes drive the observed diurnal cycles of tropical deep convection over land and ocean, in line with previous research. This complicates the task of properly capturing the diurnal cycles of deep convection in reanalysis products, km-scale or in highly parameterised ESMs. Here, novel satellite products like CCIC with a high temporal resolution will help to identify and assess biases of the modelled diurnal cycles of deep convection in the tropics and to better understand the underlying drivers of deep convection.