Diabatic heating of mesoscale convective cloud systems from synergistic satellite data

Upper tropospheric clouds are most abundant in the tropics and often form as cirrus anvils from convective outflow, building mesoscale systems (MCS). While latent heating is released into the atmosphere by the precipitating parts of these MCSs, the long-lasting anvils play a crucial role in modulating the Earth’s energy budget and heat transport. Convective organization may change the relationship between latent and radiative heating within the MCSs.

We present a coherent long-term dataset which describes tropical UT cloud systems for process and climate studies. In order to investigate also the cirrus surrounding these anvils, we used CIRS (Clouds from IR Sounders) data, retrieved from AIRS (Atmospheric InfraRed Sounder) and IASI (Infrared Atmospheric Sounding Inferometer) measurements, together with atmospheric and surface properties from the meteorological ERA reanalyses as input to artificial neural network (ANN) models to simulate the cloud vertical structure and radiative heating rates derived from CloudSat radar – CALIPSO lidar measurements, available only along narrow nadir tracks. In this way, we could expand this sparse sampling in space and in time. Furthermore, a rain rate classification, with an accuracy of about 70%, allows us to build objects of strong precipitation to identify convective organization. This dataset is now available at https://gewex-utcc-proes.aeris-data.fr/data/.

We could demonstrate that this rain intensity classification is more efficient than cold brightness temperatures to detect large latent heating, the latter derived from radar measurements of the Tropical Rainfall Measuring Mission (TRMM). While TRMM provides a diurnal sampling over a month, the spatial coverage within a time window of one hour is only about 7%. Therefore, we also expanded these latent heating profiles over the whole tropics, using ANN regression. The zonal averages of vertically integrated latent heating (LP) align well with those from the full diurnal sampling of TRMM–SLH over ocean.

In combination with a cloud system analysis we found that deeper convection leads to larger heavy rain areas, with a slightly smaller thick anvil emissivity. Convective organization enhances the mean atmospheric cloud radiative effect (ACRE) of the MCSs, in particular at small rain intensity. The projection of different MCS properties in the LP-ACRE plane can be further used for a process-oriented evaluation of parameterizations in climate models.