Learning about the ocean carbon cycle from observational constraints and model simulations of multiple tracers

A key question in studies of the potential for reducing uncertainty in climate change projections is how additional observations may be used to constrain models. We examine the case of ocean carbon cycle models. The reliability of ocean models in projecting oceanic CO₂ uptake is fundamentally dependent on their skills in simulating ocean circulation and air–sea gas exchange. In this study we demonstrate how a model simulation of multiple tracers and utilization of a variety of observational data help us to obtain additional information about the parameterization of ocean circulation and air–sea gas exchange, relative to approaches that use only a single tracer. The benefit of using multiple tracers is based on the fact that individual tracer holds unique information with regard to ocean mixing, circulation, and air–sea gas exchange. In a previous modeling study, we have shown that the simulation of radiocarbon enables us to identify the importance of parameterizing sub-grid scale ocean mixing processes in terms of diffusive mixing along constant density surface (isopycnal mixing) and the inclusion of the effect of mesoscale eddies. In this study we show that the simulation of phosphate, a major macronutrient in the ocean, helps us to detect a weak isopycnal mixing in the upper ocean that does not show up in the radiocarbon simulation. We also show that the simulation of chlorofluorocarbons (CFCs) reveals excessive upwelling in the Southern Ocean, which is also not apparent in radiocarbon simulations. Furthermore, the updated ocean inventory data of man-made radiocarbon produced by nuclear tests (bomb ¹⁴C) enable us to recalibrate the rate of air–sea gas exchange. The progressive modifications made in the model based on the simulation of additional tracers and utilization of updated observational data overall improve the model’s ability to simulate ocean circulation and air–sea gas exchange, particularly in the Southern Ocean, and has great consequence for projected CO₂ uptake. Simulated global ocean uptake of anthropogenic CO₂ from pre-industrial time to the present day by both previous and updated models are within the range of observational-based estimates, but with substantial regional difference, especially in the Southern Ocean. By year 2100, the updated model estimated CO₂ uptake are 531 and 133 PgC (1PgC?=?10¹⁵ gram carbon) for the global and Southern Ocean respectively, whereas the previous version model estimated values are 540 and 190 PgC.     

Momentum Transfer and Turbulent Kinetic Energy Budgets within a Dense Model Canopy

Second-order closure models for the canopy sublayer (CSL) employ aset of closure schemes developed for `free-air' flow equations andthen add extra terms to account for canopy related processes. Muchof the current research thrust in CSL closure has focused on thesecanopy modifications. Instead of offering new closure formulationshere, we propose a new mixing length model that accounts for basicenergetic modes within the CSL. Detailed flume experiments withcylindrical rods in dense arrays to represent a rigid canopy areconducted to test the closure model. We show that when this lengthscale model is combined with standard second-order closureschemes, first and second moments, triple velocity correlations,the mean turbulent kinetic energy dissipation rate, and the wakeproduction are all well reproduced within the CSL provided thedrag coefficient (C_D) is well parameterized. The maintheoretical novelty here is the analytical linkage betweengradient-diffusion closure schemes for the triple velocitycorrelation and non-local momentum transfer via cumulant expansionmethods. We showed that second-order closure models reproducereasonably well the relative importance of ejections and sweeps onmomentum transfer despite their local closure approximations.Hence, it is demonstrated that for simple canopy morphology (e.g.,cylindrical rods) with well-defined length scales, standard closureschemes can reproduce key flow statistics without much revision.When all these results are taken together, it appears that thepredictive skills of second-order closure models are not limitedby closure formulations; rather, they are limited by our abilityto independently connect the drag coefficient and the effectivemixing length to the canopy roughness density. With rapidadvancements in laser altimetry, the canopy roughness densitydistribution will become available for many terrestrialecosystems. Quantifying the sheltering effect, the homogeneity andisotropy of the drag coefficient, and more importantly, thecanonical mixing length, for such variable roughness density isstill lacking.     

Improvement of the K-profile Model for the Planetary Boundary Layer based on Large Eddy Simulation Data

Modifications of the widely used K-profile model of the planetary boundary layer (PBL), reported by Troen and Mahrt (TM) in 1986, are proposed and their effects examined by comparison with large eddy simulation (LES) data. The modifications involve three parts. First, the heat flux from the entrainment at the inversion layer is incorporated into the heat and momentum profiles, and it is used to predict the growth of the PBL directly. Second, profiles of the velocity scale and the Prandtl number in the PBL are proposed, in contrast to the constant values used in the TM model. Finally, non-local mixing of momentum was included. The results from the new PBL model and the original TM model are compared with LES data. The TM model was found to give too high PBL heights in the PBL with strong shear, and too low heights for the convection-dominated PBL, which causes unrealistic heat flux profiles. The new PBL model improves the predictability of the PBL height and produces profiles that are more realistic. Moreover, the new PBL model produces more realistic profiles of potential temperature and velocity. We also investigated how each of these three modifications affects the results, and found that explicit representation of the entrainment rate is the most critical.     

An Analytical Model for Mean Wind Profiles in Sparse Canopies

Existing analytical models for mean wind profiles within canopies are applicable only in dense canopy scenarios, where all momentum is absorbed by canopy elements and, hence, the effect of the ground on turbulent mixing is not important. Here, we propose a new analytical model that can simulate mean wind profiles within sparse canopies under neutral conditions. The model adopts a linearized canopy-drag parametrization and a first-order turbulence closure scheme taking into account the effects of both the ground and canopy elements on turbulent mixing. The resulting wind profile within a sparser canopy appears to be more like a logarithmic form, with the no-slip condition at the ground being satisfied. The analytical solution converges exactly to the standard surface-layer logarithmic wind profile in the case of zero canopy density (i.e., no-canopy scenario) and tends to be an exponential wind profile for a dense canopy; this feature is unique compared with existing analytical models for canopy wind profiles. Results from the new model are in good agreement with those from laboratory experiments and numerical simulations.     

Buoyancy reversal, decoupling and the transition from stratocumulus to shallow cumulus topped marine boundary layers

The transition in a marine boundary layer (MBL) from stratocumulus topped to shallow cumulus topped is investigated by using a large eddy simulation (LES) model. The experiments performed aim to examine the influence on the transition of (1) the probability of buoyancy reversal at the MBL top (i.e. situations in which the mixture of two air parcels becomes denser than either of the original parcels due to phase change or other nonlinear processes involved in the mixing), and (2) the degree of decoupling in the MBL (i.e. the strength of a shallow stably stratified layer near cloud base). Our results suggest that a stratocumulus-topped MBL is most likely to transit to a cumulus-topped one when (1) there exists high probability of buoyancy reversal at the MBL top, and (2) the MBL is decoupled due to large surface evaporation. We argue that a parameterization that includes representation of those two effects combined has the potential to provide a simple way of predicting the MBL transition in climate models.     

On the scavenging of SO2 by large and small rain drops: V. A wind tunnel and theoretical study of the desorption of SO2 from water drops containing S(IV)

An experimental and theoretical study has been carried out to investigate the rate of desorption of SO₂ from water drops falling at terminal velocity in air. The experiments were carried out in the Mainz vertical wind tunnel in which water drops of various sizes containing S(IV) in various concentrations were freely suspended in the vertical airstream of the tunnel. The results of these experiments were compared with the predictions of three theoretical models, and with the experiments of Walceket al. This comparison shows that the predictions of the diffusion model of Kronig and Brink in the formulation given by Walcek and Pruppacher agree well with the experimental results for all relevant large and small rain-drop sizes, and for all considered concentrations of S(IV) inside the drops. In contrast, the predictions of the diffusion model which assumes complete internal mixing inside a drop agrees with the experimental results only if the concentration of S(IV) inside the drop is less than that equivalent of an equilibrium SO₂ concentration of 15 ppbv. At larger concentrations, the theoretical predictions of the model for complete internal mixing progressively deviate from the experimental results. It is further shown that Barrie's double film model can be used to interpret the resistance to diffusion inside a drop in terms of a diffusion boundary layer inside the drop which increases in thickness with decreasing concentration of S(IV). Applying our results to the desorption of SO₂ from small and large rain drops falling below an assumed cloud base, shows that for typical contents of S(IV) inside the drops substantial amounts of SO₂ will desorb from these drops unless H₂O₂ is present in the surrounding air.     

The Third GABLS Intercomparison Case for Evaluation Studies of Boundary-Layer Models. Part B: Results and Process Understanding

We describe and analyze the results of the third global energy and water cycle experiment atmospheric boundary layer Study intercomparison and evaluation study for single-column models. Each of the nineteen participating models was operated with its own physics package, including land-surface, radiation and turbulent mixing schemes, for a full diurnal cycle selected from the Cabauw observatory archive. By carefully prescribing the temporal evolution of the forcings on the vertical column, the models could be evaluated against observations. We focus on the gross features of the stable boundary layer (SBL), such as the onset of evening momentum decoupling, the 2-m minimum temperature, the evolution of the inertial oscillation and the morning transition. New process diagrams are introduced to interpret the variety of model results and the relative importance of processes in the SBL; the diagrams include the results of a number of sensitivity runs performed with one of the models. The models are characterized in terms of thermal coupling to the soil, longwave radiation and turbulent mixing. It is shown that differences in longwave radiation schemes among the models have only a small effect on the simulations; however, there are significant variations in downward radiation due to different boundary-layer profiles of temperature and humidity. The differences in modelled thermal coupling to the land surface are large and explain most of the variations in 2-m air temperature and longwave incoming radiation among models. Models with strong turbulent mixing overestimate the boundary-layer height, underestimate the wind speed at 200 m, and give a relatively large downward sensible heat flux. The result is that 2-m air temperature is relatively insensitive to turbulent mixing intensity. Evening transition times spread 1.5 h around the observed time of transition, with later transitions for models with coarse resolution. Time of onset in the morning transition spreads 2 h around the observed transition time. With this case, the morning transition appeared to be difficult to study, no relation could be found between the studied processes, and the variation in the time of the morning transition among the models.     

Sensitivity of the Parameterizations of Vertical Mixing and Radiative Heat Fluxes on the Seasonal Evolution of the Thermal Structure of Lake Erie

A sensitivity analysis of the parameterizations of vertical mixing and radiative fluxes on the seasonal evolution of Lake Erie's thermal structure is performed using the Massachusetts Institute of Technology general circulation model (MITgcm) and the General Estuarine Transport Model (GETM). The models have the same horizontal resolution and are forced with observed meteorological data from April to October of 2002 and 2008. For turbid waters like Lake Erie, the three-band model for the parameterization of downward shortwave radiation produces more accurate temperatures in the thermocline and less error in simulating the mixed-layer depths than the widely used two-band model. Although the two models differ in vertical and horizontal mixing, numerical methods, and vertical discretization, they produced qualitatively comparable results. Comparison with observations shows that the models can reproduce the time evolution of the lake temperature reasonably well. The MITgcm and the GETM with the Mellor-Yamada level 2.5 (MY2.5) closure produce a deeper mixed layer than observed at a station located in the eastern basin, causing large errors in simulating the temperature in the thermocline while the GETM, using a turbulence scheme called “gen,” reproduces a mixed layer in better agreement with observations. The mixed-layer obtained with the k-ε closure is between those obtained with gen and MY2.5. The error in simulating the mixed-layer depths and the thermocline temperature at a station located in the central basin using the gen closure and the GETM was about 2°C lower than that obtained by the K-Profile Parameterization mixing scheme of the MITgcm. The models simulated a lake-wide anticyclonic circulation occupying the southwest part of the central basin but showed distinct differences in simulating gyres in the northwestern part of the central basin and in the eastern basin of the lake. The signature of a basin-scale Poincaré wave observed in the current data is also well represented by the two models.     

The Influence Of Vertical Wind Direction Shear On Dispersion In The Convective Boundary Layer, And Its Incorporation In Coastal Fumigation Models

The mean concentration distributionwithin a plume released from a point source in the atmosphericboundary layer can be greatly influenced by the systematic turningof wind with height (i.e. vertical wind direction shear). Such aninfluence includes a deflection of the plume centroid, with anassociated shearing of the vertical plume cross-section, and anenhancement of dispersion, in the horizontal plane. Wind directionshear is normally not accounted for in coastal fumigation models,although dispersion observations with shear acting as acontrolling parameter are not uncommon. A three-dimensionalLagrangian stochastic model is used to investigate the influenceof uniform wind direction shear on the diffusion of a point-sourceplume within the horizontally homogeneous convective boundarylayer, with the source located at the top of the boundary layer.Parameterisations are developed for the plume deflection andenhanced dispersion due to shear within the framework of aprobability density function (PDF) approach, and compared with theLagrangian model results. These parameterisations are thenincorporated into two applied coastal fumigation models: a PDFmodel, and a commonly used model that assumes uniform andinstantaneous mixing in the vertical direction. The PDF modelrepresents the vertical mixing process more realistically. A moreefficient version of the PDF model, which assumes a well-mixedconcentration distribution in the vertical at large times, isapplied to simulate sulfur dioxide data from the Kwinana CoastalFumigation Study. A comparison between the model results and thedata show that the model performs much better when the wind-sheareffects are included.     

Fundamental Framework and Experiments of the Third Generation of IAP/LASG World Ocean General Circulation Model

1.IntroductionThefirstbaroclinicoceanicgeneralcirculationmodel(OGCM)developedattheStateKeyLaboratoryofNumericalModelingforAtmosphericSciencesandGeophysicalFluidDynamics(LASG),InstituteofAtmosphericPhysics(IAP)isafour--layermodelwithitshorizontalresol...     

Impact of vertical mixing induced by small vertical scale structures above and within the equatorial thermocline on the tropical Pacific in a CGCM

Oceanic vertical mixing is known to influence the state of the equatorial ocean which affects the climate system, including the amplitude of El Niño/Southern Oscillation (ENSO). Recent measurements of ocean currents at high vertical resolution capture numerous small vertical scale structures (SVSs) within and above the equatorial thermocline that contribute significantly to vertical mixing but which are not sufficiently resolved by coarse resolution ocean models. We investigate the impact of the vertical mixing induced by the SVSs on the mean state and interannual variability in the tropical Pacific by using a coupled general circulation model. The vertical mixing induced by the SVSs is represented as an elevated vertical diffusivity from the surface down to the 20 °C isotherm depth, a proxy for the depth of the thermocline. We investigate different forms for the elevated mixing. It is found that the SVS-induced mixing strongly affect the mean state of the ocean leading to a warming of sea surface temperature (SST) and associated deepening and sharpening of the thermocline in the eastern equatorial Pacific. We find that the SST warming induced by the elevated mixing is further strengthened through the Bjerknes feedback and SST-shortwave flux feedback. We also find a reduction in the number of large amplitude ENSO events and in certain cases an increase in the skewness of ENSO.     

A LIMITED-LENGTH-SCALE k-ε MODEL FOR THE NEUTRAL AND STABLY-STRATIFIED ATMOSPHERIC BOUNDARY LAYER

Analytical and numerical models of the neutral and stably-stratifiedatmospheric boundary layer are reviewed. Theoretical arguments andcomputational models suggest that a quasi-steady state is attainable in aboundary layer cooled from below and it is shown how this may be incorporatedwithin a time-steady, one-dimensional model. A new length-scale-limitedk- model is proposed for flows where a global maximum mixing length isimposed by the finite boundary-layer depth or, in stably-stratifiedconditions, by the Obukhov length, whilst still reducing to a form consistentwith the logarithmic law in the surface layer. Simulations compare favourablywith data from the Leipzig experiment and from Cardington airfield inEngland.     

The effects of mixing and spreading on density in near-field river plumes

Water mass modification in surface-trapped, near-field river plumes is examined using a 1.5-layer reduced gravity model and a three-dimensional numerical model. Solutions to the layer model are shown to be qualitatively similar to previous observations and three-dimensional simulations of near-field plumes. Analytic analysis of the layer model demonstrates how the near-field plume is controlled by the competing processes of mixing and spreading. The two models are then used to explore the parameter space dependence of density changes within the near-field plume and their associated cross-shore length scales. Both the magnitude of density changes and their length scales are proportional to either estuarine discharge or fresh water discharge; density changes are also inversely proportional to the estuary mouth width. One surprising feature of the parameter space solutions is that the density of water exiting the near-field plume, a measure of the net dilution of the entire near-field plume, is shown to be inversely proportional to local mixing rates. This is because when local mixing is lower, the influence of plume spreading becomes greater; this spreading accelerates the plume, requiring more net mixing to bring the plume back to subcritical flow.     

A New Triple-Moment Blowing Snow Model

This paper presents a new triple-moment blowing snow model PIEKTUK-T by including predictive equations for three moments of the gamma size distribution. Specifically, predictive equations for the total number concentration, total mass mixing ratio, and total radar reflectivity for blowing snow are included. Tests in the context of idealized experiments and observed case studies demonstrate that the triple-moment model performs better than the double-moment model PIEKTUK-D in predicting the evolution of the number concentration, mixing ratio, shape parameter, and visibility in blowing snow, provided that the fall velocities for the total number concentration, mass mixing ratio, and radar reflectivity are weighted by the same order of the respective moments in both models. The power law relationship between the radar reflectivity factor and particle extinction coefficient found in PIEKTUK-T is consistent with one observed in snow storms. Coupling of the triple-moment blowing snow model to an atmospheric model would allow realistic studies of the effect of blowing snow on weather and climate.     

Fundamental framework and experiments of the third generation of IAP / LASG world ocean general circulation model

A new generation of the IAP / LASG world ocean general circulation model is designed and presented based on the previous 20-layer model, with enhanced spatial resolutions and improved parameterizations. The model uses a triangular-truncated spectral horizontal grid system with its zonal wave number of 63 (T63) to match its atmospheric counterpart of a T63 spectral atmosphere general circulation model in a planned coupled ocean-atmosphere system. There are 30 layers in vertical direction, of which 20 layers are located above 1000 m for better depicting the permanent thermocline. As previous ocean models developed in IAP / LASG, a free surface (rather than “rigid-lid” approximation) is included in this model. Compared with the 20-layer model, some more detailed physical parameterizations are considered, including the along / cross isopycnal mixing scheme adapted from the Gent-MacWilliams scheme. The model is spun up from a motionless state. Initial conditions for temperature and salinity are taken from the three-dimensional distributions of Levitus’ annual mean observation. A preliminary analysis of the first 1000-year integration of a control experiment shows some encouraging improvements compared with the twenty-layer model, particularly in the simulations of permanent thermocline, thermohaline circu?lation, meridional heat transport, etc. resulted mainly from using the isopycnal mixing scheme. However, the use of isopycnal mixing scheme does not significantly improve the simulated equatorial thermocline. A series of numerical experiments show that the most important contribution to the improvement of equatori?al thermocline and the associated equatorial under current comes from reducing horizontal viscosity in the equatorial regions. It is found that reducing the horizontal viscosity in the equatorial Atlantic Ocean may slightly weaken the overturning rate of North Atlantic Deep Water.     

Stability dependent parameterisations of vertical mixing in ocean general circulation models

Summary Parameterisations of mixing induced through shear instability, internal wave breaking, and double diffusion are investigated in simulations of ocean climate using a global ocean general circulation model (OGCM). Focus is placed on the sensitivity of the large scale circulation, water mass formation and transport of heat as measures of the model's ability to represent current climate. The model resolution is typical of OGCMs being coupled to atmospheric. GCMs in climate models and the parameterisations investigated are all computationally inexpensive enough to allow for integrations on long time scales. Under the assumption of constant vertical eddy coefficients (the control case), the model climatology displays acceptable values of North Atlantic Deep Water formation, Antarctic Circumpolar Current (ACC) transport, and Indonesian through-flow but an excessively deep and diffuse pycnocline structure with weak stratification in the deep ocean. It is found that various circulation and water mass properties are sensitive to the choice of parameterisation of vertical mixing and that determining a scheme which works satisfactorily over all regions (tropical, mid-latitude, and polar) of the domain is not straightforward. Parameterisations of internal wave breaking or upper ocean shear instability lead to some improvements in the model water mass formation. ACC and poleward heat transport when compared to the control case whereas parameterisations of double diffusive processes did not. Based on these and other results, various recommendations are made for mixing parameterisations in ocean climate models.With 8 Figures     

Upgrades to the Boundary-Layer Scheme in the Met Office Numerical Weather Prediction Model

Recent upgrades to the boundary-layer scheme in the UK Met Office operational global Numerical Weather Prediction model are documented. These comprise a reduction in turbulent mixing in stable conditions over the sea, and the inclusion of non-local momentum mixing in convective conditions. The dependence of low-level winds on changing stability is shown to have been significantly improved. Crucially, it is also found that these improvements in local performance have been achieved without degrading the model skill in terms of synoptic evolution—something that has proved difficult to achieve in the past in many operational models. In fact some aspects of the large-scale flow (e.g. zonal mean winds) have been slightly improved.     

Nonlocal Turbulent Mixing Based on Convective Adjustment Concepts (Ntac)

A nonlocal turbulent mixing parameterization is introduced in this study and denoted by the acronym NTAC, which stands for Nonlocal parameterization of Turbulent mixing using convective Adjustment Concepts. NTAC uses the average value of quantities in the turbulent domain in much the same way that local convective adjustment schemes use the average potential temperature. Averages are determined in the region with non-convective turbulence using information from the two end layers (denoted by TLA, Two Layer Average), while all layers contribute to the average in regions with convective turbulence (denoted by CLA, Convective Layer Average). The NTAC parameterization estimates the mixing percentage and uses this percentage as a mixing coefficient. These percentages are determined from a simplified turbulent kinetic energy equation. The scheme is versatile, conservative, and when programmed efficiently the proposed parameterization is a computationally acceptable nonlocal procedure that can be used in many existing numerical weather prediction forecast models.Numerical weather forecast model simulations using the NTAC parameterization and traditional K-theory are compared against radiosonde data. The accuracy of the proposed NTAC parameterization is found to be competitive with K theory. The greatest improvement of the NTAC over K-theory occurs during the daytime and early nighttime hours when (dry) convective activity is high. Also, areal cloud coverage is increased by the NTAC parameterization. Our findings show that the greatest nonlocal vertical mixing occurs between the layer nearest the earth's surface and the remaining layers making up the planetary boundary layer.     

Effects of turbulence on gas-phase atmospheric chemistry: Calculation of the relationship between time scales for diffusion and chemical reaction

Summary Non-uniform mixing of gas-phase trace species may limit the accuracy of the predictions of Eulerian transport/transformation models if the chemical reactions are rapid enough to be diffusion limited. If a reaction is diffusion limited, its average reaction rate might not be accurately represented by those models that assume instantaneous uniform mixing. One possible consequence of this artificial dilution is the overprediction of ozone and hydroxyl radicals. We have determined which reactions in the Regional Acid Deposition Model Gas-Phase Chemical Mechanism (Stockwell et al., 1990) are diffusion limited for a typical atmospheric condition through the calculation of Damköhler numbers. Damköhler numbers are defined to be the ratio of the diffusion mixing time to the chemical reaction time for a given chemical reaction (McRae et al., 1982; Hill, 1976). The reactions of hydroxyl radicals and the reactions of peroxy radicals with NO are diffusion limited under typical atmospheric conditions. Both sets of reactions are especially significant because NO_x and organic species strongly affect ozone and hydroxyl radical concentrations. It is suggested that Damköhler numbers could be used to help determine the placement of Eulerian model boundaries and to determine model grid structure.With 2 Figures