Gas exchange rates measured using a dual-tracer (SF6 and3he) method in the coastal waters of Korea

Over a period of 5 days between August 12 and 17, 2005, we performed a gas exchange experiment using the dual tracer method in a tidal coastal ocean located off the southern coast of Korea. The gas exchange rate was determined from temporal changes in the ratio of³He to SF₆ measured daily in the surface mixed layer. The measured gas exchange rate (k _{CO 2}), normalized to a Schmidt number of 600 for CO₂ in fresh water at 20°C, was approximately 5.0 cm h^-1 at a mean wind speed of 3.9 m s^-1 during the study period. This value is significantly less than those obtained from floating chamber-based experiments performed previously in estuarine environments, but is similar in magnitude to values obtained using the dual tracer method in river and tidal coastal waters and values predicted on the basis of the relationship between the gas exchange rate and wind speed (Wanninkhof 1992), which is generally applicable to the open ocean. Our result is also consistent with the relationship of Raymond and Cole (2001), which was derived from experiments carried out in estuarine environments using²²²Rn and chlorofluorocarbons along with measurements undertaken in the Hudson River, Canada, using SF⁶ and³He. Our results indicate that tidal action in a microtidal region did not discernibly enhance the measuredk _{CO 2} value.     

The effect of statistical wind corrections on global wave forecasts

The skill of modern wave models is such that the quality of their forecasts is, to a large degree, determined by errors in the forcing wind field. This work explores the extent to which large-scale systematic biases in modelled waves from a third generation wave model can be attributed to the forcing winds. Three different sets of winds with known global bias characteristics are used to force the WAVEWATCH III model. These winds are based on the Australian Bureau of Meteorology’s ACCESS model output, with different statistical corrections applied. Wave forecasts are verified using satellite altimeter data. It is found that a negative bias in modelled Significant Wave Height (H_s) has its origins primarily in the forcing, however, the reduction of systematic wind biases does not result in universal improvement in modelled H_s. A positive bias is present in the Southern Hemisphere due primarily to an overestimation of high H_s values in the Southern Ocean storm tracks. A positive bias is also present in the east Pacific and East Indian Ocean. This is due both to the over-prediction of waves in the Southern Ocean and lack of swell attenuation in the wave model source terms used. Smaller scale features are apparent, such as a positive bias off the Cape of Good Hope, and a negative bias off Cape Horn. In some situations, internal wave model error has been compensated for by error in the forcing winds.     

Inter-Comparison of Numerical Model Generated Surface Winds with QuikSCAT Winds over the Indian Ocean

The accurate surface wind in the equatorial Indian Ocean is crucial for modeling ocean circulation over this region. In this study, the surface wind analysis generated at the European Center for Medium Range Weather Forecasts (ECMWF) and the National Centers for Environmental Prediction (NCEP) are compared with NASA QuikSCAT satellite derived Level2B (swath level) and Level3 (gridded) surface winds for the year 2005. It is observed that the ECMWF winds exhibit speed bias of 1.5 m/s with respect to QuikSCAT Level3 in the southern equatorial Indian Ocean. The NCEP winds are found to exhibit speed bias (1.0–1.5 m/s) in the southern equatorial Indian Ocean specifically during January–February 2005. The biases are also observed in the analysis when compared with Level2B product as well; however, it is less in comparison to Level3 products. The amplitude of daily variations of both ECMWF and NCEP wind speed in Bay of Bengal and parts of the Arabian Sea is about 80% of that in QuikSCAT, while in the equatorial Indian Ocean it is about 60% of that of QuikSCAT.     

Roles of Physical Processes in the Carbon Cycle Using a Simplified Physical Model

A simplified physical model is proposed in this article to describe differences among basins in substance distributions which were not well described by previous simplified models. In the proposed model, the global ocean is divided into the Pacific/Indian Ocean (PI), the Atlantic Ocean (AT), the Southern Ocean and the Greenland/Iceland/Norwegian Sea. The model is consisted of five physical parameters, namely the air-sea gas exchange, the thermohaline circulation, the horizontal and vertical diffusions, and the deep convection in the high-latitude regions. Individual values of these parameters are chosen by optimizing model distribution of natural ¹⁴C as a physical tracer. The optimal value for a coefficient of vertical diffusion in the low-latitude region is 7.5 × 10^–5 [m²s^–1]. Vertical transports by the Antarctic Bottom Water and the North Atlantic Deep Water are estimated at 1.0 Sv and 9.0 Sv. Global-mean air-sea gas exchange time is calculated at 9.0 years. Using these optimal values, vertical profiles of dissolved inorganic carbon without biological production in PI and AT are estimated. Oceanic responses to anthropogenic fluctuations in substance concentrations in the atmosphere induced by the industrialization and nuclear bomb are also discribed, i.e., the effects appear significantly in AT while a signal is extremely weak in PI. A time-delay term is effective to make the PI water older near the bottom boundary.     

Measurements of the oceanic thermal skin effect

Spectroradiometric measurements of the ocean skin temperature and thermometric measurements of the bulk temperature at a depth of about 5 cm taken from the R/V Tangaroa during SAGE (SOLAS/SAGE: surface-ocean lower-atmosphere studies air-sea gas exchange experiment) off New Zealand are analyzed to reveal the wind speed dependence of the temperature difference across the thermal skin layer (??T). The wind speeds used here are corrected for flow distortion by the ship. Unlike most previously published measurements of ??T, these data include those taken during the day, prior analyses being usually restricted to night-time measurements to avoid contamination of the data by diurnal heating. The results show the same dependence of ??T on wind speed at night-time measurements, with an asymptotic behavior at a value of −0.13 K at high winds. These data show larger ??T at low wind speeds than previous studies, and there is an indication that this may reveal a dependence on sea surface temperature.     

Equatorial influence of QuikSCAT winds in an isopycnal ocean model compared to NCEP2 winds

An ocean model was used to examine whether the scatterometer winds can improve the model performance both dynamically and thermodynamically. Comparisons were done using QuikSCAT and NCEP2 winds for both the mean and variability from 2000 to 2004. The comparisons showed that the model forced by QuikSCAT winds gives more realistic mean SST, 20 °C isotherm depth (Z20), and latent heat flux than NCEP2 winds do. Sensitivity experiments indicated that QuikSCAT mean wind stress is important for the improved mean SST, Z20, and latent heat release to the atmosphere in the eastern Pacific. QuikSCAT wind speed, through its effect on the turbulent heat fluxes, is most important for the mean SST in the western Pacific. Finally, there were comparable correlations with observations of both SST and Z20 on the intra-seasonal time scale between the model forced with QuikSCAT winds and the model forced with NCEP2 winds.     

Evaluation of relative performance of QuikSCAT and NCEP re-analysis winds through simulations by an OGCM

The response of an ocean general circulation model (OGCM) to two different wind products, viz., NCEP/NCAR reanalysis and QuikSCAT scatterometer, was examined. OGCM-simulated thermodynamic variables from the two simulations, hereafter referred to as NCEP-R (NCEP/NCAR wind forced) and QS-R (QuikSCAT wind forced) were intercompared and also were compared against observations for a period of 3 years (2000–2002). In the tropical Indian Ocean (IO), the sea-level anomaly (SLA) simulated by QS-R has less root mean square error (RMSE) and higher correlation with respect to TOPEX/Poseidon SLA observations than SLA simulated by NCEP-R. Intraseasonal variability of currents observed by TRITON buoy in the IO was closely captured by QS-R, although the magnitudes are somewhat underestimated. Surface currents simulated by QS-R have less RMSE than those simulated by NCEP-R in the Pacific. However, the sub-surface currents are much weaker in magnitude in both the solutions, possibly because of deficiencies in the diffusion and viscosity parameterization. Sea-surface temperature (SST) simulated by QS-R has a cooler bias. The RMSE of SST simulated by NCEP-R is less than the RMSE of SST simulated by QS-R, with the latter capturing the variabilities more realistically. The large differences between SST simulated by QS-R and observations could be partly due to physical inconsistency between the momentum and heat fluxes. Scatterometer-forced model simulations of 20^oC thermocline depths (D20) are in better agreement with in situ-derived D20 than the D20 simulated by NCEP-R. Variations in the mixed layer depth at the TRITON buoy are better captured by QS-R than by NCEP-R. Speed of Kelvin and Rossby waves and the strength of upwelling/downwelling features in the IO are closer to observations in QS-R than in NCEP-R simulations.     

生物固氮速率测定中15N2同位素示踪剂的引入

生物固氮作用是一个重要的海洋新氮来源,在海洋生物地球化学循环中扮演着不可替代的角色。基于稳定同位素（15N₂）示踪吸收法,是目前直接测定海洋生物固氮速率最有效的手段。其中,高效、洁净地将15N₂引入海水培养体系,并准确定量培养体系底物的同位素示踪水平,是同位素示踪吸收法准确获取固氮速率的关键。本研究针对15N₂同位素示踪剂引入这一关键环节进行了探讨,确认改进气泡法是将15N₂引入海水培养体系的首选操作。在对培养体系造成的较小扰动的情况下,可将培养体系氮气底物的15N原子丰度提升至10%以上,相对于另一种导入同位素示踪剂的手段——预溶解海水法,改进气泡法将培养瓶中氮气底物的15N原子丰度提升了近200%。此外,改进气泡法还具有最小化痕量金属沾污、操作简便等优点。将改进气泡法结合与稳定同位素比值质谱测定结合,是准确测定水体生物固氮速率的推荐方法。     

The SOLAS air-sea gas exchange experiment (SAGE) 2004

The SOLAS air-sea gas exchange experiment (SAGE) was a multiple-objective study investigating gas-transfer processes and the influence of iron fertilisation on biologically driven gas exchange in high-nitrate low-silicic acid low-chlorophyll (HNLSiLC) Sub-Antarctic waters characteristic of the expansive subpolar zone of the southern oceans. This paper provides a general introduction and summary of the main experimental findings. The release site was selected from a pre-voyage desktop study of environmental parameters to be in the south-west Bounty Trough (46.5°S 172.5°E) to the south-east of New Zealand and the experiment was conducted between mid-March and mid-April 2004. In common with other mesoscale iron addition experiments (FeAX’s), SAGE was designed as a Lagrangian study, quantifying key biological and physical drivers influencing the air-sea gas exchange processes of CO₂, DMS and other biogenic gases associated with an iron-induced phytoplankton bloom. A dual tracer SF₆/³He release enabled quantification of both the lateral evolution of a labelled volume (patch) of ocean and the air-sea tracer exchange at tenths of kilometer scale, in conjunction with the iron fertilisation. Estimates from the dual-tracer experiment found a quadratic dependency of the gas exchange coefficient on windspeed that is widely applicable and describe air-sea gas exchange in strong wind regimes. Within the patch, local and micrometeorological gas exchange process studies (100 m scale) and physical variables such as near-surface turbulence, temperature microstructure at the interface, wave properties and windspeed were quantified to further assist the development of gas exchange models for high-wind environments.There was a significant increase in the photosynthetic competence (F_v/F_m) of resident phytoplankton within the first day following iron addition, but in contrast to other FeAX’s, rates of net primary production and column-integrated chlorophyll a concentrations had only doubled relative to the unfertilised surrounding waters by the end of the experiment. After 15 days and four iron additions totalling 1.1 ton Fe²⁺, this was a very modest response compared to other mesoscale iron enrichment experiments. An investigation of the factors limiting bloom development considered co-limitation by light and other nutrients, the phytoplankton seed-stock and grazing regulation. Whilst incident light levels and the initial Si:N ratio were the lowest recorded in all FeAXs to date, there was only a small seed-stock of diatoms (less than 1% of biomass) and the main response to iron addition was by the picophytoplankton. A high rate of dilution of the fertilised patch relative to phytoplankton growth rate, the greater than expected depth of the surface mixed layer and microzooplankton grazing were all considered as factors that prevented significant biomass accumulation. In line with the limited response, the enhanced biological draw-down of pCO₂ was small and masked by a general increase in pCO₂ due to mixing with higher pCO₂ waters. The DMS precursor DMSP was kept in check through grazing activity and in contrast to most FeAX’s dissolved dimethylsulfide (DMS) concentration declined through the experiment. SAGE is an important low-end member in the range of responses to iron addition in FeAX’s. In the context of iron fertilisation as a geoengineering tool for atmospheric CO₂ removal, SAGE has clearly demonstrated that a significant proportion of the low iron ocean may not produce a phytoplankton bloom in response to iron addition.     

Did dilution limit the phytoplankton response to iron addition in HNLCLSi sub-Antarctic waters during the SAGE experiment?

An in situ iron addition experiment (SAGE) was carried out in high-nitrate low-chlorophyll low-silicic acid (HNLCLSi) sub-Antarctic surface waters south-east of New Zealand. In contrast to other iron addition experiments, the phytoplankton response was minor, with a doubling of biomass relative to surrounding waters, with the temporal trends in dissolved iron and macronutrients instead dominated by physical factors such as mixing and dilution. The initial increase in patch surface area indicated a lateral dilution rate of 0.125 d⁻¹, with a second estimate from a model of the decline in peak SF₆ concentration yielding a higher lateral dilution rate of 0.16-0.25 d⁻¹. The model was tested on the SOIREE SF₆ dataset and provided a lateral dilution of 0.07 d⁻¹, consistent with previous published estimates. MODIS ocean colour images showed elevated chlorophyll coincident with the SF₆ patch on day 10 and 12, and an elevated chlorophyll filament at the SAGE experiment location 3-4 days after ship departure, which provided additional lateral dilution estimates of 0.19 and 0.128 d⁻¹. Dissolved iron at the patch centre declined by 85% within two days of the initial infusion, of which dilution accounted for 50-65%; it also decreased rapidly after the 2nd and 3rd infusions but remained elevated after the fourth infusion. Despite decreases in nitrate and silicic acid from day 7 and 10, respectively, the final nutrient concentrations in the patch exceeded the initial concentrations due to supply from lateral intrusion and mixed-layer deepening. The low Si:N loss ratio suggested that the observed limited response to iron was primarily by non-siliceous phytoplankton. Algal growth rate exceeded the minimum dilution rate during two periods (days 3-6 and 10-14), and coincided with net chlorophyll accumulation. However, as the ratio of algal growth to dilution was the lowest reported for an iron addition experiment, dilution was clearly a significant factor in the SAGE experiment recording the lowest phytoplankton response to mesoscale iron addition.     

Grazing suppression of dimethylsulphoniopropionate (DMSP) accumulation in iron-fertilised, sub-Antarctic waters

The impact of in situ iron fertilisation on the production of particulate dimethylsulphoniopropionate (DMSPp) and its breakdown product dimethyl sulphide (DMS) was monitored during the SOLAS air-sea gas exchange experiment (SAGE). The experiment was conducted in the high nitrate, low chlorophyll (HNLC) waters of the sub-Antarctic Southern Ocean (46.7°S 172.5°E) to the south-east of New Zealand, during March-April, 2004. In addition to monitoring net changes in the standing stocks of DMSPp and DMS, a series of dilution experiments were used to determine the DMSPp production and consumption rates in relation to increased iron availability. In contrast to previous experiments in the Southern Ocean, DMS concentrations decreased over the course of the 15-d iron-fertilisation experiment, from an integrated volume-specific concentration in the mixed layer on day 0 of 0.78 nM (measured values 0.65-0.91 nM) to 0.46 nM (measured values 0.42-0.47 nM) by day 15, in parallel with the surrounding waters. DMSPp, chlorophyll a and the abundance of photosynthetic picoeukaryotes exhibited indiscernible or only moderate increases in response to the raised iron availability, despite an obvious physiological response by the phytoplankton. High specific growth rates of DMSPp, equivalent to 0.8-1.2 doublings d⁻¹, occurred at the simulated 60% light level of the dilution experiments. Despite the high production rates, DMSPp accumulation was suppressed in part by microzooplankton grazers who consumed between 61% d⁻¹ and 126% d⁻¹ of the DMSPp production. Temporal trends in the rates of production and consumption illustrated a close coupling between the DMSP-producing phytoplankton and their microzooplankton grazers. Similar grazing and production rates were observed for the eukaryotic picophytoplankton that dominated the phytoplankton biomass, partial evidence that picoeukaryotes contributed a substantial proportion of the DMSP synthesis. These rates for DMSPp and picoeukaryotes were considerably higher than for chlorophyll a, indicating higher cycling rates of the DMSP-producing taxa than for the bulk phytoplankton community. When compared to the total phytoplankton community, there was no evidence of selection against the DMSP-containing phytoplankton by the microzooplankton grazers; the opposite appeared to be the case. SAGE demonstrated that increased iron availability in the HNLC waters of the Southern Ocean does not invariably lead to enhanced DMS sea-air flux. The potential suppression of DMSPp accumulation by grazers needs to be taken into account in future attempts to elevate DMS emission through in situ iron fertilisation and in understanding the hypothesised link between levels of Aeolian iron deposition in the Southern Ocean, DMS emission and global albedo.     

The Rossby wave as a key mechanism of Indian Ocean climate variability

We analyze the time-longitude structure of composite cases from model-assimilated ocean data in the period 1958–1998, following on from earlier work by Huang and Kinter (J. Geophys. Res. 107(C11) (2002) 3199) that studied east–west thermocline variability in the Indian Ocean. Our analysis focuses on the Rossby wave signal along the thermocline ridge in the tropical SW Indian Ocean (10°S, 60–80°E), where wind stress curl is important. Anomalous winds in the equatorial east Indian Ocean force successive Rossby waves westward at speeds of 0.1 m s⁻¹±30%. With a wavelength of 7000 km, the period of oscillation is in the range 1.9–5.2 years. The Indian Ocean Rossby wave is partially resonant with the global influence of the El Nino–Southern Oscillation, except during quasi-biennial rhythm. The presence of the Rossby wave offers potential predictability for east–west atmospheric circulation systems and climate that affect resources in countries surrounding the Indian Ocean.     

Carbon dioxide flux techniques performed during GasEx-98

A comprehensive study of air–sea interactions focused on improving the quantification of CO₂ fluxes and gas transfer velocities was performed within a large open ocean CO₂ sink region in the North Atlantic. This study, GasEx-98, included shipboard measurements of direct covariance CO₂ fluxes, atmospheric CO₂ profiles, atmospheric DMS profiles, water column mass balances of CO₂, and measurements of deliberate SF₆–³He tracers, along with air–sea momentum, heat, and water vapor fluxes. The large air–sea differences in partial pressure of CO₂ caused by a springtime algal bloom provided high signals for accurate CO₂ flux measurements. Measurements were performed over a wind speed range of 1–16 m s⁻¹ during the three-week process study. This first comparison between the novel air-side and more conventional water column measurements of air–sea gas transfer show a general agreement between independent air–sea gas flux techniques. These new advances in open ocean air–sea gas flux measurements demonstrate the progress in the ability to quantify air–sea CO₂ fluxes on short time scales. This capability will help improve the understanding of processes controlling the air–sea fluxes, which in turn will improve our ability to make regional and global CO₂ flux estimates.     

Air-sea gas transfer in a shallow,flowing and coastal environment estimated by dissolved inorganic carbon and dissolved oxygen analyses

We estimated gas exchange rates in Kabira Reef at Ishigaki Island, southwest Japan, using a mass balance calculation with dual “biological” tracers: dissolved inorganic carbon (DIC) and dissolved oxygen (DO). The nighttime results allowed us to obtain reasonable gas transfer velocity k _w values, all of which exceeded those obtained in wind-dominant studies. The difference is likely due to the contribution of turbulence generated by the interaction between the current and bottom topography. The k _w obtained during high tides is consistent with that reported by Raymond and Cole (2001), whereas k _w during low tides is significantly higher, which seems to be caused by enhanced friction with the bottom of the reef and/or bubble-induced gas transfer by wave breaking at the reef crest.     

Processes governing the carbon chemistry during the SAGE experiment

Measurements of pCO₂, pH and alkalinity in the surface waters of an iron fertilised patch of sub-Antarctic water were made during SAGE (SOLAS SAGE: Surface-Ocean Lower Atmosphere Studies Air-Sea Gas Experiment). The iron addition induced a minor phytoplankton bloom, however the patch dynamics were dominated by physical processes which suppressed and masked the biological effects. The Lagrangian nature of the experiment allowed the carbonate chemistry in the patch to be followed for 15.5 days, and the relative importance of the biological and physical factors influencing the surface water pCO₂ was estimated. The pCO₂ of the surface waters of the patch increased from 327 ??atm prior to iron addition to 338 ??atm on Day 14, effects of vertical and horizontal mixing offset the 15 ??atm drawdown that would have occurred had the induced biological uptake been the sole factor to influence the pCO₂. The air-sea carbon flux calculated using the measured skin temperature and a piston velocity parameterisation determined during SAGE (Ho et al., 2006) was 98.5% of the flux determined using conventional bulk temperature measurement and the Wanninkhof (1992) piston velocity parameterisation. The skin temperature alone contributed to an 8% increase in the flux compared with that determined using bulk temperature.     

Tuning a physically-based model of the air–sea gas transfer velocity

Air–sea gas transfer velocities are estimated for one year using a 1-D upper-ocean model (GOTM) and a modified version of the NOAA–COARE transfer velocity parameterization. Tuning parameters are evaluated with the aim of bringing the physically based NOAA–COARE parameterization in line with current estimates, based on simple wind-speed dependent models derived from bomb-radiocarbon inventories and deliberate tracer release experiments. We suggest that A = 1.3 and B = 1.0, for the sub-layer scaling parameter and the bubble mediated exchange, respectively, are consistent with the global average CO₂ transfer velocity k. Using these parameters and a simple 2nd order polynomial approximation, with respect to wind speed, we estimate a global annual average k for CO₂ of 16.4 ± 5.6 cm h^?1 when using global mean winds of 6.89 m s^?1 from the NCEP/NCAR Reanalysis 1 1954–2000. The tuned model can be used to predict the transfer velocity of any gas, with appropriate treatment of the dependence on molecular properties including the strong solubility dependence of bubble-mediated transfer. For example, an initial estimate of the global average transfer velocity of DMS (a relatively soluble gas) is only 11.9 cm h^?1 whilst for less soluble methane the estimate is 18.0 cm h^?1.     

Natural air–sea flux of CO2 in simulations of the NASA-GISS climate model: Sensitivity to the physical ocean model formulation

Results from twin control simulations of the preindustrial CO₂ gas exchange (natural flux of CO₂) between the ocean and the atmosphere are presented here using the NASA-GISS climate model, in which the same atmospheric component (modelE2) is coupled to two different ocean models, the Russell ocean model and HYCOM. Both incarnations of the GISS climate model are also coupled to the same ocean biogeochemistry module (NOBM) which estimates prognostic distributions for biotic and abiotic fields that influence the air–sea flux of CO₂. Model intercomparison is carried out at equilibrium conditions and model differences are contrasted with biases from present day climatologies. Although the models agree on the spatial patterns of the air–sea flux of CO₂, they disagree on the strength of the North Atlantic and Southern Ocean sinks mainly because of kinematic (winds) and chemistry (pCO₂) differences rather than thermodynamic (SST) ones. Biology/chemistry dissimilarities in the models stem from the different parameterizations of advective and diffusive processes, such as overturning, mixing and horizontal tracer advection and to a lesser degree from parameterizations of biogeochemical processes such as gravitational settling and sinking. The global meridional overturning circulation illustrates much of the different behavior of the biological pump in the two models, together with differences in mixed layer depth which are responsible for different SST, DIC and nutrient distributions in the two models and consequently different atmospheric feedbacks (in the wind, net heat and freshwater fluxes into the ocean).     

Surface layer mixing during the SAGE ocean fertilization experiment

Vessel-based observations of the oceanic surface layer during the 14-day 2004 SAGE ocean fertilization experiment were conducted using ADCP, CTD and temperature microstructure in a frame of reference moving with a patch of injected SF₆ tracer. During the experiment the mixed layer depth z_mld ranged between 50 and 80 m, with several re-stratifying events that brought z_mld up to less than 40 m. These re-stratifying events were not directly attributable to local surface-down development of stratification and were more likely associated with horizontal variation in density structure. Comparison between the CTD and a one-dimensional model confirmed that the SAGE experiment was governed by 3-d processes. A new method for estimating z_mld was developed that incorporates a component that is proportional to density gradient. This highlighted the need for well-conditioned near-surface data which are not always available from vessel-based survey CTD profiles. A centred-displacement scale, L_c, equivalent to the Thorpe lengthscale, reached a maximum of 20 m, with the eddy-centroid located at around 40 m depth. Temperature gradient microstructure-derived estimates of the vertical turbulent eddy diffusivity of scalar (temperature) material yielded bin-averaged values around 10⁻³ m² s⁻¹ in the pycnocline rising to over 10⁻² m² s⁻¹ higher in the surface layer. This suggests transport rates of nitrate and silicate at the base of the surface layer generate mixed layer increases of the order of 38 and 13 mmol/m²/day, respectively, during SAGE. However, the variability in measured vertical transport processes highlights the importance of transient events like wind mixing and horizontal intrusions.