| Popis: |
Shallow trade-wind cumulus clouds have been a research focus in the recent decade due to their uncertain response to climate change. While in some models the occurrence of these clouds is strongly reduced under future climate conditions, it remains nearly invariant in others. These contradictory results reflect our limited understanding of the mechanisms involved in the formation and spatial organisation of these clouds. Processes on different scales are at play: turbulent fluxes, convective mixing, large-scale subsidence, and the clouds themselves, which affect the thermodynamical conditions in their surroundings through diabatic processes (i.e., cloud radiative effects, warming through condensation, evaporative cooling). To shed light on the complex cloud-circulation coupling, the field campaign Elucidate the Couplings Between Clouds, Convection and Circulation (in short EUREC4A) was carried out in 2020 in the vicinity of the Caribbean island Barbados. The intensive observation period took place from 20 January to 20 February. The overarching goal of the campaign was to provide observational constraints for these low-level cloud processes and their interaction with atmospheric circulations. Already before EUREC4A, it was known that these clouds frequently organise themselves in different mesoscale patterns referred to as Sugar, Gravel, Flower, and Fish. These mesoscale cloud organisation patterns were readily adopted by the EUREC4A community to classify the cloudiness during the campaign. The campaign included a component called EUREC4Aiso, which operated a stable water isotopologue measurement network consisting of two aircrafts, four ships, and the Barbados Cloud Observatory (BCO), which was the major island-based observational site of EUREC4A. Stable water isotopologues (hereafter water isotopes or isotopes) are non-radioactive water molecules with different molecular structures and weights due to a heavy hydrogen (2H) or oxygen (18O) atom. Here, the isotopes 1H216O, 1H2H16O, and 1H218O were considered and the isotope parameter d2H and d-excess were used to assess isotope effects. The above-mentioned isotopes have different saturation vapour pressures and diffusion velocities, which means that their relative abundance in a given water sample changes during phase transition. The isotopic composition, thus, reflects the accumulated record of all moist atmospheric processes (e.g., evaporation, condensation and mixing) that the sample has experienced since evaporation from the ocean (which is the reference used for the normalisation of atmospheric isotope ratios). Thus, isotopes provide a link to the Lagrangian history of atmospheric water and have the ability to link processes on different scales. The measurements conducted by the author include measurements of water vapour isotopes in the lower troposphere onboard the French aircraft ATR-42 during 19 research flights that took place between 25 January and 13 February 2020, the continuous measurements of water vapour isotopes at the BCO from 13 January to 17 February 2020, and the collection of 42 rain samples at the BCO from 16 January to 18 February 2020. High consistency was found between the isotope measurements from the different EUREC4Aiso platforms except for an unexpected offset between the two instruments deployed for the vapour measurements at the BCO. Overall, the comparison of the isotope measurements from the different platforms showed that our measurements are of high quality and suitable for model evaluations and process-oriented analyses. The EUREC4A observations were complemented with ERA5 reanalyses and numerical simulations performed with the isotope-enabled, regional model COSMOiso in a setting with fully resolved convection. Applying a spectral nudging, nested simulations with spatial resolutions of 10 km, 5 km, 1km (referred to as COSMOiso,10km, COSMOiso,5km, COSMOiso,1km), were performed. The simulated cloud organisation patterns approximate the ones found in satellite images. Fish, Flower, and finer-structured (i.e., Gravel and Sugar) cloud patterns are clearly distinguishable in the simulations. However, cold pools, which are closely connected to the different patterns, are only evident in the COSMOiso,1km data. Precipitation is simulated in smaller regions compared to satellite observations. Cloud fraction as well as liquid water contents at cloud base are larger than in aircraft observations. These variables are strongly dependent on the spatial model resolution and generally, they approximate the observations better the higher the resolution. Contrastingly, temperature, humidity, and isotopes in vapour are nearly insensitive to the spatial model resolution. While their temporal variability (except for temperature) matches well with observations, deviations were found with respect to absolute values. The simulations are characterized by a low-level cold-dry bias associated with too strongly depleted vapour and an overly high d-excess. Despite these shortcomings, the high correlation between simulated and measured time series at the synoptic timescale showed that the simulations represent the large-scale forcing in an adequate way. Furthermore, a feature-based analysis of the isotope variability at cloud base showed that the key processes of the mesoscale moisture cycling are well captured by the model. The role of the large-scale circulation for the trade-wind region and its cloudiness was addressed using trajectories that reach 10 d (using ERA5 winds) or 6 d (with winds from COSMOiso,10km) backward in time. They showed that the air parcels arriving in the lower troposphere near Barbados typically take one of three pathways. Either they cross the North Atlantic at low latitudes with little vertical displacement (typical trade-wind flow) or they descend from high latitudes and altitudes within an extratropical dry intrusion in the rear of a Rossby wave breaking. If the Rossby wave breaking occurs over the eastern North Atlantic/Europe, the air parcels cross the North Atlantic (similarly to the typical trade-wind flow) after completing their descent. If the Rossby waver breaking is situated over the central/western North Atlantic, the air parcels arrive much more directly after their descent (extratropical dry intrusion flow). The different pathways are reflected in the subsidence rate of the trajectories, which is weakest for air parcels following a tropical trade-wind flow and strongest for air parcels travelling within an extratropical dry intrusion flow regime. A link could be established between the subsidence rate and the isotopic composition in the trade-wind region: the stronger the subsidence, the lower the d2H and the higher the d-excess of the vapour and precipitation. In other words, flow regime-specific processes lead to contrasting isotope signals and, therefore, isotopes can be used to identify changes in the large-scale circulation. The extratropical dry intrusion flow regime was shown to play a crucial role for the occurrence of the Fish cloud pattern. The dry intrusion air parcels either arrive together with the trailing cold front leading to moist and cloudy conditions or they arrive shortly afterwards in the cold sector leading to dry and clear-sky conditions. Independently of this, on a daily basis, the arrival of dry intrusion air parcels in Barbados generally leads to a radiative cooling through the shortwave effect of the Fish cloud and the longwave effect of the dryness in the free troposphere (caused by the strengthened large-scale subsidence within the dry intrusion). Another process that is induced by the large-scale circulation and alters the radiative balance, although not directly connected to shallow clouds, is the formation of an altostratus cloud layer in the mid troposphere. Such an event occurred towards the end of EUREC4A. The trajectory analysis showed that this mid-level cloud layer was formed due to detrainment of moist air parcels from tropical convection at the 0°C-isotherm and subsequent northward transport towards Barbados. The search for the origin of the isotopic contrasts between cloudy and clear-sky cloud base environments led to novel insights into the overturning circulation associated with clouds. A link was found between the isotopic characteristics of clear-sky environments and the altitude at which detrainment from shallow clouds takes place. Moreover, the comparison of the COSMOiso,1km data to a simple model simulating fractionation due to condensation in a moist adiabatically cooled air parcel during ascent revealed that vapour in COSMOiso,1km is enriched through evaporation in the lower part of the cloud layer and depleted through mixing with surrounding vapour in the upper part of the cloud layer. Finally, strong up- and downdrafts in clouds (assumed to represent shallow convective mixing) were shown to moisten and enrich the lower troposphere more than weak shallow convective mixing. The work carried out in the framework of this thesis clearly reveals the prominent role of the large-scale circulation for the trade-wind region and its cloudiness. Furthermore, it shows that isotopes reflect changes in the large-scale circulation and provide constraints for processes embedded in the overturning circulation associated with clouds. By adding a complementary perspective to conventional meteorological variables, isotopes substantially contribute to the understanding of atmospheric processes. Combining the isotope signal and the trajectory of an air parcel, one can identify couplings across large distances, which would be difficult to diagnose otherwise. |
