Periodic Reporting for period 4 - COAT (Collapse Of Atmospheric Turbulence)
Período documentado: 2020-07-01 hasta 2020-12-31
This project aims to predict the cessation of continuous turbulence in the evening boundary layer.
The interaction between the lower atmosphere and the surface is studied in detail, as this plays a
crucial role in the dynamics. Present generation forecasting models are incapable to predict whether
or not turbulence will survive or collapse under cold conditions. In nature, both situations frequently
occur and lead to completely different temperature signatures. As such, significant forecast errors
are made, particularly in arctic regions and winter conditions. Therefore, prediction of turbulence
collapse is highly relevant for weather and climate prediction.
Key innovation lies in our hypothesis. The collapse of turbulence is explained from a maximum
sustainable heat flux hypothesis which foresees in an enforcing positive feedback between the
atmosphere and the underlying surface. A comprehensive theory for the transition between the main
two nocturnal regimes would be ground-breaking in meteorological literature.
We propose an integrated approach, which combines in-depth theoretical work, simulation with
models of various hierarchy (DNS, LES, RANS), and observational analysis. Such comprehensive
methodology is new with respect to the problem at hand. An innovative element is the usage of
Direct Numerical Simulation in combination with dynamical surface interactions. This advanced
technique fully resolves turbulent motions up to their smallest scale without the need to rely on
subgrid closure assumptions. From a 10-year dataset (200m mast at Cabauw, Netherlands) nights
are classified according to their turbulence characteristics.
Multi-night composites are used as benchmark-cases to guide realistic numerical modelling. In the
validation phase, generality of the results with respect to both climate and surface characteristics is
assessed by comparison with the FLUXNET data-consortium, which operates on a long-term basis
over 240 sites across the globe.
The interaction between the lower atmosphere and the surface is studied in detail, as this plays a
crucial role in the dynamics. Present generation forecasting models are incapable to predict whether
or not turbulence will survive or collapse under cold conditions. In nature, both situations frequently
occur and lead to completely different temperature signatures. As such, significant forecast errors
are made, particularly in arctic regions and winter conditions. Therefore, prediction of turbulence
collapse is highly relevant for weather and climate prediction.
Key innovation lies in our hypothesis. The collapse of turbulence is explained from a maximum
sustainable heat flux hypothesis which foresees in an enforcing positive feedback between the
atmosphere and the underlying surface. A comprehensive theory for the transition between the main
two nocturnal regimes would be ground-breaking in meteorological literature.
We propose an integrated approach, which combines in-depth theoretical work, simulation with
models of various hierarchy (DNS, LES, RANS), and observational analysis. Such comprehensive
methodology is new with respect to the problem at hand. An innovative element is the usage of
Direct Numerical Simulation in combination with dynamical surface interactions. This advanced
technique fully resolves turbulent motions up to their smallest scale without the need to rely on
subgrid closure assumptions. From a 10-year dataset (200m mast at Cabauw, Netherlands) nights
are classified according to their turbulence characteristics.
Multi-night composites are used as benchmark-cases to guide realistic numerical modelling. In the
validation phase, generality of the results with respect to both climate and surface characteristics is
assessed by comparison with the FLUXNET data-consortium, which operates on a long-term basis
over 240 sites across the globe.
Goal of the project is to “predict of the collapse of turbulence in the evening boundary layer and to understand the nonlinear coupling between the atmosphere and its underlying surface.” The plan of the proposal was followed in phase I: a 10-year dataset from Cabauw was used to classify nocturnal boundary layer regimes (and their degree of turbulence) in terms of the external forcings (large-scale pressure gradient, degree of cloud coverage). The resulting composites, published in Van der Linden et al. (2017) serve as a benchmark for numerical modeling. In Baas et al. (2017) we assessed the state-of-the-art of practical modeling with respect to this classification, using RACMO (basically the IFS scheme of the European Centre for Medium-range Weather Forecast). In the next stage, the impact of regime transition on fog formation was assessed.
Also, in accordance with the plan, we were successful in extending the maximum sustainable heat flux theory with land-atmosphere interactions, by introducing a new conceptual model (Van de Wiel et al., 2017; Figures below). As to contrast results from The Netherlands, also data from Antarctica (i.e. with snow coverage instead of grass) where analyzed with respect to collapsing turbulence and regime transitions in (Vignon et. al. 2017), and a comparison with the conceptual model was given in Van de Wiel et al. (2017). In the theoretical part of the project, simplified cases without atmospheric-surface interactions where studies using turbulence resolving techniques such as Direct Numerical Simulation in order to understand and predict the collapse of turbulence in those configurations. Successful ‘early warnings of critical transitions’ where reported in Van Hooijdonk et al. (2016). As, in reality, the collapse of turbulence is influenced by the turbulent state of the atmosphere during the preceding day and in the evening, we also investigated the impact of the evening transition in Van Hooijdonk et al. (2017) and Van Hooft et al. (2019). As stated in the proposal special emphasis has been given to conceptualization of land-atmosphere interactions and a generalized description of the diurnal cycle of the dry atmospheric boundary layer was given in Van Hooft et al. (2019), where indeed it was shown that (nighttime) boundary layer response was very different for different forcings.
To enable turbulence resolving simulations of the full diurnal cycle, we also investigated the potential of adaptive-grids in LES/DNS (Van Hooft et al., 2017). This technique appears to be promising in highly dynamic problems that suffer from a large-scale separation, such as our problem at hand. According to the research plan model results have been benchmarked against results reported in the literature. In summary: the project was executed according to the plan and significant progress was obtained, which has been disseminated in the scientific literature.
Also, in accordance with the plan, we were successful in extending the maximum sustainable heat flux theory with land-atmosphere interactions, by introducing a new conceptual model (Van de Wiel et al., 2017; Figures below). As to contrast results from The Netherlands, also data from Antarctica (i.e. with snow coverage instead of grass) where analyzed with respect to collapsing turbulence and regime transitions in (Vignon et. al. 2017), and a comparison with the conceptual model was given in Van de Wiel et al. (2017). In the theoretical part of the project, simplified cases without atmospheric-surface interactions where studies using turbulence resolving techniques such as Direct Numerical Simulation in order to understand and predict the collapse of turbulence in those configurations. Successful ‘early warnings of critical transitions’ where reported in Van Hooijdonk et al. (2016). As, in reality, the collapse of turbulence is influenced by the turbulent state of the atmosphere during the preceding day and in the evening, we also investigated the impact of the evening transition in Van Hooijdonk et al. (2017) and Van Hooft et al. (2019). As stated in the proposal special emphasis has been given to conceptualization of land-atmosphere interactions and a generalized description of the diurnal cycle of the dry atmospheric boundary layer was given in Van Hooft et al. (2019), where indeed it was shown that (nighttime) boundary layer response was very different for different forcings.
To enable turbulence resolving simulations of the full diurnal cycle, we also investigated the potential of adaptive-grids in LES/DNS (Van Hooft et al., 2017). This technique appears to be promising in highly dynamic problems that suffer from a large-scale separation, such as our problem at hand. According to the research plan model results have been benchmarked against results reported in the literature. In summary: the project was executed according to the plan and significant progress was obtained, which has been disseminated in the scientific literature.
We primarily followed the path foreseen in the plan: theoretical modeling was augmented with fully interactive surface schemes and besides turbulence also radiative effects were accounted for.
This enables comparison with atmospheric observational data. Also the 'snow-covered' case was investigated as for such case, strongest regime transitions are expected. Apart from modeling with advanced turbulence models, also simulations with state-of-the-art practical models were made (e.g. Baas et al. 2017, 2019). Also the effect of collapsing turbulence on fog formation was explored in depth, with special emphasis on analysis of local surface characteristics and their influence on appearance disappearance of fog (Izett et al. 2018, 2019). Finally (as stated above): to enable turbulence resolving simulations of the full diurnal cycle, we investigated the potential of adaptive-grids in LES/DNS (Van Hooft et al., 2017). This technique appears to be promising in highly dynamic problems that suffer from a large-scale separation, such as our problem at hand. In summary: we made significant progress in the understanding of the cessation of turbulence in the evening boundary layer and also in understanding its practical implications.
This enables comparison with atmospheric observational data. Also the 'snow-covered' case was investigated as for such case, strongest regime transitions are expected. Apart from modeling with advanced turbulence models, also simulations with state-of-the-art practical models were made (e.g. Baas et al. 2017, 2019). Also the effect of collapsing turbulence on fog formation was explored in depth, with special emphasis on analysis of local surface characteristics and their influence on appearance disappearance of fog (Izett et al. 2018, 2019). Finally (as stated above): to enable turbulence resolving simulations of the full diurnal cycle, we investigated the potential of adaptive-grids in LES/DNS (Van Hooft et al., 2017). This technique appears to be promising in highly dynamic problems that suffer from a large-scale separation, such as our problem at hand. In summary: we made significant progress in the understanding of the cessation of turbulence in the evening boundary layer and also in understanding its practical implications.