The contribution of salt fingers to vertical mixing in the North Atlantic Tracer Release Experiment

Louis St Laurent, MIT

4:30pm, Thursday, 30 Jan 1997

Abstract The North Atlantic Tracer Release Experiment (Ledwell, Watson and Law; 1993) was performed in an area moderately favorable to salt fingers. However, the classic finger signature of a distinct thermohaline staircase caused by the up-gradient buoyancy flux was absent. This is likely because mixing by turbulence is sufficiently strong to prevent the formation of sharp interfaces. However, optical shadowgraph profiles revealed that the tilted laminae previously observed in a salt finger staircase (Kunze et al, 1987) were a dominant feature of the microstructure at the NATRE site. A more quantitative test for fingers was made by examining the apparent efficiency of the mixing processes producing the microscale velocity and thermal microstructure as measured by 150 dives of the "High Resolution Profiler" (HRP, Schmitt et al, 1988). This was done by evaluating the "Scaled dissipation ratio" (Gamma) introduced by Oakey (1988). Because the HRP provides a complete suit of finescale velocity and density measurements in addition to the microstructure, the variation of Gamma can be examined in a parameter space defined by the density ratio and the Richardson number. We find that low Richardson number regions have a dissipation ratio consistent with previous models and data for turbulence, regardless of the density ratio. However, for Richardson numbers above 1/2, a distinct elevation of Gamma into the range predicted for salt fingers is observed for density ratios less than 2. The region local to the tracer release surface (sigma-theta = 26.75, roughly 300 m depth) is characterized by low density ratio with Richardson numbers near 1. We estimate that 20% to 40% of the diapycnal flux of tracer can be attributed to salt fingers.

Infragravity wave pressure and velocity variance in the nearshore

Tom Lippmann (Scripps)

4:30pm, Thursday, 6 Feb 1997

Keywords Infragravity, edge waves, leaky waves, instabilities of the longshore current

Abstract An across-shore array of 9 co-located pressure sensors and bi-directional current meters, extending from the shoreline to ~4.5 m depth, is used to determine the cross-shore variation of both infragravity (frequencies 0.004-0.05 Hz) surface gravity wave (edge and leaky) and low frequency shear wave (linear instabilities of the mean longshore current) pressure and velocity variance. Under the assumption that amplitudes are independent of wavenumber (broad along dispersion curves), theoretical predictions on a plane beach show that the integrated cross-shore (longshore) velocity to pressure variance ratio is approximately g/2h (g/2h) for directionally- broad low mode edge waves, and g/h (0) for directionally-narrow high mode edge waves and leaky waves, where g is gravity and h is the local water depth. Predicted ratios for shear waves are >> g/h (>> g/h). The observed total velocity to pressure variance ratios (including both cross-shore and alongshore components), normalized by g/2h, is 2 for surface gravity waves and >> 2 for shear waves. Outside the surf zone where gradients in the longshore current profile are small (compared to inside the surf zone), observed normalized variance ratios (integrated over the infragravity frequencies and divided by g/2h) are approximately equal to 2, consistent with an infragravity spectrum composed of primarily free edge and leaky waves. Inside the surf zone the ratios are much larger, as much as 8 on some days, indicating that shear waves contribute as much as 75% of the total velocity variance. Energetic shear waves are confined to the (often) narrow region of strong seaward shear in the mean longshore current, and have spatial variation which does not depend directly on the local cross-shore bathymetric profile, consistent with the simple theoretical predictions of Bowen and Holman (1989).

The use of deep bores to obtain large scale measurements of evaporation, rainfall and groundwater flow

Steve Henderson, Dalhousie

4:30pm, Thursday, 13 Feb 1997

Keywords Evaporation estimation, rainfall estimation, groundwater recharge

Abstract The water level in a deep bore varies with atmospheric pressure, earth tides and changes in the weight of material above the base of the bore. Weight changes are normally dominated by changes in water storage, hence it has been suggested that it may be possible to measure rainfall, evaporation, groundwater recharge, divergent groundwater flow etc. on large scales using bore water level records (van der Kamp and Maathuis, 1991; Bardsley and Campbell, 1994). Water storage estimates obtained from two closely spaced bores are presented. Good agreement is found between aquifer-based and independent estimates of water storage change. A method is developed for the identification and removal of earth tidal effects. After the effects of earth tides are removed a diurnal fluctuation remains in the estimated water storage time series. This fluctuation may be caused by evaporation, however it appears that systematic measurement errors may have contaminated the signal.

A general pressure gradient scheme and its applications to North Atlantic modelling.

Yuhe T. (Tony) Song and Dan Wright BIO/Dalhousie

4:30pm, Thursday, 20 Feb 1997

Keywords modelling

Abstract A new formulation of the pressure gradient force for use in models with topography-following coordinates is proposed. It is based on a Jacobian formulation, it can be used in conjunction with any vertical coordinate system, and it is easily implemented.

Two discrete schemes are derived: the first using standard centered differencing in the computational vertical coordinate and the second using vertical weighting such that the hydrostatic inconsistency is eliminated for density perturbation fields which vary quadratically. Both schemes achieve second order accuracy for any vertical coordinate and are significantly more accurate than the conventional scheme based on applying finite differencing directly to the pressure field.

The new schemes are numerically consistent, energetically consistent and accurately represent the bottom pressure torque. Their performances in a large-scale wind-driven basin are tested prognostically for both numerical accuracy and long-term integral stability, based on a model with and without topography. The integrations are carried out for 10 years in each case and results show that the schemes are stable, and the steep topography causes no obvious problems. A realistic meandering western boundary current is well developed with detached cold cyclonic and warm anticyclonic eddies as it extends across the basin. In addition, the results with topography show earlier separation and enhanced transport in the western boundary currents due to the bottom pressure torque.

A North Atlantic model with the new pressure gradient schemes has been developed for the Canadian Community Modeling Effort. The model domain is from 7N to 71N and from 98W to 15E with 1/3 degree resolution.

A Case Study of an Extratropical Storm South of Nova Scotia

Chris Fogarty

4:30pm, Thursday, 6 Mar 1997

Abstract On November 14th, 1996 a low pressure system formed 300 miles south of Nova Scotia and moved quickly to the east. By Saturday November 16th the low was deepening quickly and began to curve to the north and eventually northwest around 42N/58W. After drifting slowly southward, the low became fully occluded and began to weaken late on the 16th. A second deepening occurred on the 17th as the storm moved south over 24-degree Gulf Stream waters. Accompanying the re-intensification were convective bands of rainfall and thunderstorms north and west of a clear central eye seen on satellite images. Slow weakening followed, however, convective bands continued to appear on satellite imagery with an intermittent eye. This tropical like behavior continued until late on the 18th. The storm slowly drifted eastward thereafter, still over warm Gulf Stream waters, and eventually sped off to the northeast.

With use of ship, buoy, and station data as well as FEM model outputs and satellite images of the period, the question of whether the system was becoming tropical in nature, is answered.

The behavior of the low after its "extratropical maturity" is what has been analyzed. Normally a cut-off low in the midlatitudes (as such) dies after the 500-mb minimum in geopotential heights align with the surface pressure minimum below. The low is said to "fill" and the surface pressure rises. However, with this particular case, the re-intensification of the surface low after it became "vertically stacked" cannot be explained by traditional midlatitude dynamic theory.

Nonlinear interactions in the atmospheric gravity-wave spectrum: a raytracing approach

Len Sonmor (Dalhousie) and Gary Klaassen (York)

4:30pm, Thursday, 20 Mar 1997

Abstract Middle-atmosphere research has followed the lead of oceanography in recognizing the importance of nonlinear interactions within the gravity-wave spectrum. Dissipative processes accompanying such interactions are thought to play an important role in mesospheric wave drag, and in the establishment of observed spectra. Eikonal (raytracing) theory is well suited for modeling the effect, on a test wave, of advective nonlinearity due to the presence of larger scale and lower frequency waves. Colin Hines' Doppler-spread theory models the gravity-wave wind component as an ensemble of time-invariant profiles that vary only in the vertical; his frozen vertical eikonal model has led to a computationally efficient wave-drag parameterization scheme that is currently used in large-scale middle-atmosphere models in several different countries. The effects of the temporal and horizontal variations that are present in the gravity-wave portion of the atmospheric winds remain largely unexplored. These have been investigated in an oceanic context [eg Broutman GAFD 1984; Flatte et al. JGR 1985; Bruhwiler and Kaper JFM 1995], and found to yield significant, and statistically non-cancelling, effects. It is unclear how to apply these results to the middle atmosphere, with its much larger mean shear, density diminution, and predominantly descending wave phases. We examine effects of relaxing the frozen vertical approximation, using a background that is more representative of atmospheric winds than the idealized and oceanic backgrounds used in earlier studies. The resulting interactions differ fundamentally from those found in a frozen background, and defy action-conservation laws that are sometimes assumed to apply.

Cloud Microphysics and Precipitation Parameterization within Large Scale Models

George Isaac

4:30pm, Thursday, 27 Mar 1997

Abstract It is difficult to simulate clouds and precipitation within large scale models used in weather forecasting and climate change studies. Many assumptions must be made within the models to determine cloud amount, cloud microphysics, cloud radiative properties, and precipitation occurrence and amount. Often, limited measurements of highly variable parameters made on a small regional spatial scale, must be used to develop parameterizations on a much larger scale over long periods of time. This is a tremendous scientific challenge. Since our weather and climate can be very sensitive to such parameterizations, it is also a very important problem to solve. This talk will present illustrations of measurements, and discuss the difficulties involved in processing such measurements to develop accurate parameterization schemes. The limitations of current schemes, especially those used in climate models, will be assessed.

Dynamics of Circulation off the west coast of Vancouver Island

Badal Pal (BIO/Dalhousie) and G. Holloway (IOS)

4:30pm, Thursday, 3 April 1997


Abstract The Princeton Ocean Model (Blumberg and Mellor, 1987) is used to examine dynamical balances in the summer and winter 3-D circulation along the west coast of Vancouver Island. The model is initialized with horizontally averaged temperature and salinity over a domain 445 km alongshore by 145 km cross-shore. Forcing is by uniform summer and winter winds and by discharge from Juan de Fuca Strait. A topographic stress parameterization (Eby and Holloway, 1994) is included.

In the absence of topographic stress and Juan de Fuca discharge, wind-forced summer circulation is dominated by equatorward flow over the shelf and slope. With topographic stress included, poleward coastal flow develops over the shelf and shelf-break with a transition zone between poleward and equatorward flow over the slope. When the buoyancy forcing is introduced, an eddy is developed off the mouth of the Juan de Fuca Strait. Winter circulation tends to be poleward for all combinations of wind, buoyancy and topographic stress forcings. These results agree qualitatively with observations.

Impact of sulfate aerosols on albedo and lifetime of clouds.

Ulrike Lohmann -- candidate for Asst. Prof. position in Atmospheric Science at Dalhousie

4:30pm, Thursday, 17 April 1997

NOTE: special location: Psychology 5260 (i.e., the room used for Tuesday seminars)

Abstract A coupled sulfur chemistry-cloud microphysics scheme (COUPL) is used to study the impact of sulfate aerosols on cloud lifetime and albedo. The cloud microphysics scheme includes precipitation formation, which depends on the cloud droplet number concentration (CDNC) and on the liquid water content. On the basis of different observational data sets, CDNC is proportional to the sulfate aerosol mass, which is calculated by the model. Cloud cover is a function of relative humidity only. Additional sensitivity experiments with another cloud cover parameterization (COUPL-CC), which also depends on cloud water, and with a different autoconversion rate of cloud droplets (COUPL-CC-Aut) are conducted to investigate the range of the indirect effect due to uncertainties in cloud physics. For each experiment, two simulations, one using present-day and one using preindustrial sulfur emissions are carried out. The increase in liquid water path, cloud cover, and shortwave cloud forcing due to anthropogenic sulfur emissions depends crucially upon the parameterization of cloud cover and autoconversion of cloud droplets. In COUPL the liquid water path increases by 17% and cloud cover increases by 1% because of anthropogenic sulfur emissions, yielding an increase in shortwave cloud forcing of -1.4 W m-2. In COUPL-CC the liquid water path increases by 32%, cloud cover increases by 3% and thus shortwave cloud forcing increases by -4.8 W m-2. This large effect is caused by the strong dependence of cloud cover on cloud water and of the autoconversion rate on CDNC, cloud water, and cloud cover. Choosing a different autoconversion rate (COUPL-CC-Aut) with a reduced dependence on CDNC and cloud water results in an increase of liquid water path by only 11% and of cloud cover by 1%, and the increase in shortwave cloud forcing amounts to -2.2 W m-2. These results clearly show that the uncertainties linked to the indirect aerosol effect are higher than was previously suggested.

Approximating the adjoint model: Variational data assimilation for operational applications using a realistic ocean model

Mark Buehner, Dalhousie

4:30pm, Thursday, 24 April 1997

Keywords ?

Abstract Our goal is to use a complex nonlinear ocean model together with observations of the ocean to produce an estimate of the ocean state at various times. Using data together with the ocean model the ocean state is estimated either within the assimilation period (hindcast), at the end of the assimilation period (nowcast), or projected beyond the data period using the model dynamics alone (forecasting). To enable the assimilation procedure to be applicable to all of these situations, a variational approach to data assimilation is followed instead of sequential methods.

Two complicating factors are soon encountered when attempting to assimilate data using realistic ocean models regardless of the method: large dimension of model state vector (typically on the order of $10^{4}$ or greater) and complex nonlinear dynamics. These have been overcome in several cases and indeed realistic ocean models have been successfully used in variational assimilation studies. However, the effort required to attain a working system and the computational expense of performing an assimilation experiment are both high. We address these difficulties by subjecting a generic ocean model to a series of approximations sufficient to make assimilation tractable while treating the model as a ``black box''. These approximations enable one to easily incorporate any ocean model into a computationally efficient assimilation procedure.

Some results using an idealized configuration of a nonlinear primitive equation model (CANDIE) are presented to demonstrate the strengths and weaknesses of the approach.

Sea surface elevations and currents over the Scotian and Newfoundland Shelves from altimetry and hydrography

Guoqi Han, Bedford Institute of Oceanography

4:30pm, Thursday, 1 May 1997

Keywords Sea surface elevation variability; shelf circulation; altimetry; hydrography; numerical modeling

Abstract The TOPEX/POSEIDON (T/P) altimeter data over the period 1992-96 have been analyzed to examine annual variability of the sea surface elevations and currents over the Scotian and Newfoundland Shelves. A regional tidal model is employed to remove the oceanic tides from the altimetric sea surface elevations over these regions. To derive the annual cycle, harmonic analysis is applied to the time series of the sea surface elevations. The sea surface slopes, which represent surface geostrophic currents, are also estimated. The altimetric results are compared with the solutions from a numerical model forced by baroclinicity, wind stress and remote forcing. The altimetric elevations at Halifax and St. John's (interpolated from nearby T/P observations) agree favorably with the tide-gauge data, with an annual amplitude of about 5 cm high in late fall and low in late spring. The comparison between the altimetric and modeled results indicates that the altimetric sea surface elevation variability is dominated by the baroclinic (and associated barotropic) component at Halifax, but contributed by both the baroclinicity and the wind-driven and remote-forced components at St. John's. The wintertime intensification of the shelf-break flows are suggested by the altimetric surface currents, consistent with the model solutions.

The Response of Canadian Atlantic Shelf Seas to Meteorological Forcing

Josko Bobanovic, Dalhousie

4:30pm, Thursday, 8 May 1997

Keywords wind driven currents; assimilation; operational modelling; Labrador Shelf; Newfoundland Shelf; Scotian Shelf; Gulf of Saint-Lawrence

Abstract As a part of an ongoing effort to develop an operational model of the circulation over the shelves of Atlantic Canada, we use a linear barotropic model to hindcast flow from Labrador Shelf to the Gulf of Maine. The model is forced by wind, air pressure and open boundary conditions. Model hindcasts are assessed by comparison against the data collected in the region. Simple statistical models are used to establish limits on the accuracy of the linear model hindcasts.

Having validated the model we examine the connections between different regions. We explore how wind driven flow on the Labrador and the Newfoundland Shelf affects the downstream shelves. It appears that the coastal trapped wave scattering on rugged topography of the Labrador Shelf prevents the signal from reaching the Gulf of Saint Lawrence. However, the wind forcing over the Newfoundland Shelf is a major source of variability for both the Gulf of Saint Lawrence and the Scotian Shelf. The connection between regions is frequency dependent. The results are used to provide a guidance for the development of data assimilative regional models.

The dynamics of jet streaks in the atmosphere

Philip Cunningham, Dept. Earth and Atm Sci., University at Albany, SUNY

4:30pm, Thursday, 15 May 1997

Keywords Jet streaks, Vortex dipoles, Analytical & Numerical Modelling, Balanced turbulence

Abstract (See here for more info on this work (with nice pictures! -- DK).

Localised along-stream maxima in the jet stream, commonly known as "jet streaks", are ubiquitous in extratropical flow and, because of their association with cyclogenesis and severe weather, have been the subject of numerous observational studies. Nevertheless, little attention has been devoted to dynamical or theoretical treatment of these features.

Motivated by the apparently prevalent association of jet streaks in the atmosphere with mesoscale coherent vortices that often occur as a dipole pair, we have examined the possibility that the theory of coherent structures, and in particular that of vortex dipoles, may provide insight into the structure, motion and evolution of jet streaks. Characteristic signatures of analytical and numerical solutions of vortex dipoles, both in isolation and in background flows representative of the extratropical upper troposphere, are examined and are shown to exhibit structure similar to that seen in observational studies of jet streaks. General conclusions regarding the interpretation of jet streaks in terms of coherent structures are discussed, and are related to some commonly accepted descriptions of jet-streak behaviour.