The effect of below-cloud processes on short-term variations of stable water isotopes in surface precipitation

Abstract

The atmospheric water cycle is a key component of our climate system and a better understanding of the involved processes is crucial to estimate the impacts of a changing climate on, e.g., the spatial distribution of extreme precipitation. In the mid-latitudes, over 80% of precipitation extremes are associated with extratropical cyclones and their occurrence or absence controls the frequency and severity of floods and droughts. Phase-change processes like condensation and rain evaporation affect the dynamics of extratropical cyclones and their impact by contributing latent heat to the cyclone intensification, and by significantly reducing the surface precipitation amount, respectively. A widely applied method to passively trace the atmospheric water cycle with a focus on phase-change processes is to consider stable water isotopes, because these naturally available tracers record the condensation and evaporation history of atmospheric water vapour and precipitation. The aim of this thesis is to identify the driving mechanisms of isotopic variability on the sub-event time scale and draw conclusions for the dynamics of the investigated events. This is done by performing parallel high-frequency measurements of stable water isotopes in near-surface vapour and precipitation for selected rain events and compare these data to simulations with a below-cloud interaction model fed with meteorological observations. In the first part of this thesis, the techniques to measure the isotopic composition of surface vapour and rain are outlined. In-situ vapour isotope measurements of δ2H, δ18O and hence d-excess are performed with two cavity ring-down laser spectrometers (Picarro L1115-i and L2130-i). Precipitation is sampled in short intervals (∼ 10 min) during several rain events and analysed for its isotopic composition. Supporting meteorological observations are taken from radiosondes, a disdrometer, a micro rain radar and two X-band radars. An overview of the data is presented, which was collected during two measurement campaigns in Switzerland. The second part introduces the below-cloud interaction model, which simulates the changing isotopic composition of a single falling hydrometeor. The model is used to enhance the conceptual understanding of below-cloud processes and test the sensitivity of the isotopic signal of rain to temperature, relative humidity, formation height, the formation mechanism, and to the isotopic composition of the source vapour. Variables that determine the initial isotopic composition of rain mostly affect large hydrometeors and intense rain. In contrast, small hydrometeors and weak rain are mostly affected by variables that determine the strength of below-cloud processes. These processes are limited by the height of the melting layer and weak if the melting layer is low. The driving mechanisms of isotopic variability of selected frontal rain events are analysed in the third part of this thesis. The evolution of the isotopic difference between rain and near-surface vapour of three cold frontal passages is compared to the mass weighted mean diameter Dm of the rain samples to identify the influence of below-cloud processes. Thereby, the importance of sampling rain with a high frequency is underlined. Phases are identified, when rain is mainly controlled by evaporation or equilibration or by changes in the source conditions. Below-cloud processes are more dominant for weak rain intensities, a high melting layer and low near-surface relative humidity. They are thus expected to have a weaker influence in cold and moist seasons and climates, where convective rain dominates. Rain in comparison with near-suface vapour is on average depleted by 5.4 permil in δ2H, which indicates incomplete equilibration, and lower by 4.6 permil in d-excess, which underlines the presence of rain evaporation. A new reference frame (∆δ∆d-space) for the interpretation of rain samples is introduced, which displays the isotopic difference between rain and near-surface vapour for δ2H and d-excess. The degree of equilibrium between rain and near-surface vapour and the influence of evaporation can be directly identified from the location of a sample in the ∆δ∆d-space. Certain periods are dominated by the influence of evaporation, and during others, the influence of below-cloud processes is weak. A characteristic slope in the ∆δ∆d-space of −0.30 ± 0.01 is found for all samples of all investigated rain events. A comparison between observations and below-cloud interaction model simulations yields that short-term, sample to sample isotopic variations are caused by changes in the drop size, and that the longer-term evolution is controlled by a change of the source vapour, which is manifested by a changing vertical isotopic profile in the model. Conclusions about the relative humidity profile can be drawn from the difference between measured and simulated d-excess. Differences of δ2H are used to constrain changes in the vapour isotopic profile due to the layering of different air masses. It is demonstrated with these analyses that stable water isotopes provide information about the dynamics of weather events that is not available from traditional meteorological variables, or that would involve disproportionately more extensive measurements. In summary, this thesis presents an unprecedented combination of parallel high-frequency measurements of stable water isotopes in vapour and rain in mid-latitude weather events. The measurements can be used to improve our understanding of below-cloud processes and of the controls of short-term variations of isotopes in precipitation. The combination with model simulations is a valuable tool to constrain atmospheric conditions and to validate atmospheric models during rain events.