3D Radiative Heating of Tropical Upper Tropospheric Cloud Systems derived from Synergistic A-Train Observations and Machine Learning
Résumé
Upper Tropospheric (UT) cloud systems constructed from Atmospheric Infrared Sounder (AIRS) cloud data provide a horizontal emissivity structure, allowing to link convective core to anvil properties. By using machine learning techniques we composed a horizontally complete picture of the radiative heating rates deduced from CALIPSO lidar and CloudSat radar measurements, which are only available along narrow nadir tracks. To train the artificial neural networks, we combined the simultaneous AIRS, CALIPSO and CloudSat data with ERA-Interim meteorological reanalysis data in the tropics over a period of four years. Resulting non-linear regression models estimate the radiative heating rates as a function of about 40 cloud, atmospheric and surface properties, with a column-integrated mean absolute error (MAE) of 0.8 K/d (0.5 K/day) for cloudy scenes and 0.4 (0.3 K/day) for clear sky in the longwave (shortwave) spectral domain. Already about 20 basic input variables yield good results, with a 6% (10%) larger MAE. Developing separate models for i) high opaque clouds, ii) cirrus, iii) mid- and low-level clouds and iv) clear sky, independently over ocean and over land, lead to a small improvement, when considering the profile shapes. These models were then applied to the whole AIRS cloud dataset, combined with ERA-Interim, to build 3D radiative heating rate fields. Over the deep tropics, UT clouds have a net radiative heating effect of about 0.3 K/day throughout the troposphere from 250 hPa downward, with a broad maximum of about 0.4 K/d around 330 hPa, enhancing the column-integrated latent heating by about 25%. This value is larger than earlier results of about 20%. Above the height of 200 hPa, the LW cooling above convective cores and thick cirrus anvils is opposed by thin cirrus heating. Whereas in cooler regions low-level clouds also influence the net radiative heating profile, in warmer regions it is nearly completely driven by deep convective cloud systems. These mesoscale convective systems (MCS) are colder and include slightly more thin cirrus around their anvils than those in cooler regions. Hence, the MCSs over these warmer regions produce a vertically more extended heating by the thicker cirrus anvils and a heating of 0.7 K/d above the height of 200 hPa by the surrounding thin cirrus. The roughly estimated horizontal gradients between cirrus anvil and convective core as well as between surrounding thin cirrus and cirrus anvil seem to be slightly smaller in warmer regions, which can be explained by their larger coverage. The 15-year time series of the heating / cooling effects of MCSs are well related to the ENSO variation. While the coverage of all MCSs is relatively stable (or very slightly decreasing) with surface warming, with -1.3 ± 0.6 %/K, the coverage of cold MCSs relative to all MCSs significantly increases by +18 ± 5 %/K.
Origine : Fichiers produits par l'(les) auteur(s)
