Workshop on the Application of EPIC2001 Data for Improving and Testing Coupled Atmosphere-Ocean Models (Abstracts)

Session 1 - Fluxes at the air-sea interface

OBSERVATIONS OF ATMOSPHERIC BOUNDARY LAYER FLOW ACROSS THE NORTH SIDE OF THE COLD TONGUE IN THE EASTERN EQUATORIAL PACIFIC

C. A. Paulson, H. Wijesekera and W. S. Pegau, Oregon State University
D. Rudnick, Scripps Institution of Oceanography

As part of the EPIC field experiment, atmospheric and oceanic observations were made in the eastern tropical Pacific from the R/V New Horizon. The atmospheric measurements included wind speed and direction, temperature, humidity, incoming short and long wave radiation, and turbulent fluxes of heat and momentum. The oceanic observations included temperature and salinity measured from the surface down to a depth of 250 m and horizontal velocity with a ship-mounted ADCP. As part of the experiment, underway observations were made at a speed of 4 m/s along 95W from 10N to 1S and back to 10N. The east-west surface temperature front at the north side of the equatorial cold tongue was located just south (0.35 deg S) of the equator and had a magnitude of 1.7 C over a distance of a 700 meters when we crossed from south to north. Surface salinity changed 0.2 psu across the front which contributed to the density change. The temperature front had a dramatic effect on the structure of the northward atmospheric boundary layer flow across the front. The wind speed at 10-m height south of the front was 4 m/s with a low level of turbulence, consistent with stable or near-neutral stratification. North of the front, wind speed gradually increased over a distance of 40 km to 9 m/s with a high level of turbulence, consistent with unstable stratification. The higher wind speed north of the front suggests that enhanced mixing, associated with unstable stratification, transports northward momentum from aloft toward the surface and that stable stratification inhibits momentum transfer south of the front. Fluxes of sensible and latent heat reached 40 and 170 W/m^2 respectively north of the front but were near zero south of the front.

Session 2 - Observations and cloud parameterizations

Session 3 - Easterly Waves / MJO

 

Session 4 - Ocean model experiments

MODELING THE EASTERN PACIFIC: THERMOCLINE ISSUES

Raghu Murtugudde, Associate Professor
ESSIC/METO, University of Maryland, College Park, MD 20742

The eastern Pacific is crucial in terms of its role in ENSO variability through Bjerknes feedback and yet the model simulations continue to be systematically biased with respect to observations. A suite of forced OGCM experiments are carried out to illustrate the factors that affect thermocline variability in the EPIC region. Higher vertical and horizontal resolutions improve model simulations but the mechanisms involved are distinct. High frequency variability in winds impacts the thermocline by enhancing mixing relative to the low-frequency forcing. The role of extra-tropical surface variability on the tropical thermocline is simulated by modifying the bottom boundary conditions and by nudging the model to observations below 300m. It is evident that improving the deeper part of the thermocline clearly improves the upper thermocline and has consequences for mixed layer-thermocline interactions. This has implications for coupled climate models and also for the role Argo data will play in improving climate models. The role of heat fluxes is illustrated by comparing the latent heat flux variability in an atmospheric mixed layer model coupled to the OGCM. Aspects of thermocline variability that is well-simulated by models and the observational requirements for further improvements are discussed.

 

SEASONAL AND INTERANNUAL VARIABILITY IN THE EASTERN PACIFIC WARM POOL

Kristopher B. Karnauskas, Antonio J. Busalacchi and Raghu Murtugudde
Earth System Science Interdisciplinary Center, University of Maryland, College Park, MD

Although the eastern Pacific warm pool (EPWP) is the warmest body of water in the eastern Pacific Ocean, relatively little research effort has been invested in understanding its variability. Infrared satellite measurements of sea surface temperature (SST), outgoing longwave radiation (OLR) and precipitation have been made for multiple decades and provide suitable records for the study of the seasonal and interannual variability of the EPWP. Furthermore, the advent of high resolution microwave satellite remote sensing and the implementation of extensive observing systems in the eastern tropical Pacific provide a means to study the finer structures within the traditional bounds of the EPWP and their own variability in detail never before possible. Preliminary analysis of such data as well as regional OGCM simulations yield provocative evidence that it is insufficient to treat the EPWP as a single, uniform body of water. Instead, thermocline topography and gap winds through the Central American cordillera appear to split the EPWP into northern, central, and southern subregions, including the Costa Rica dome, wherein each subregion exhibits its own characteristic coupled interaction as stratified by the covariability of SST and OLR. Given the importance of this region in ENSO dynamics and its impact on the climate of the Americas, further data analyses and modeling are necessary to understand the processes involved in maintaining these regional contrasts. Some early results will be presented with discussion of planned observational analyses and modeling studies.

Session 5 - Ocean eddies and fronts

UPPER OCEAN HEAT AND FRESHWATER BUDGETS IN THE EASTERN PACIFIC WARM POOL

Hemantha W. Wijesekera1, Daniel L. Rudnick2, Clayton A. Paulson1, Stephen D. Pierce1, W. Scott Pegau3, John Mickett4, and Michael C. Gregg4

1 College of Oceanic and Atmospheric Sciences, Oregon State University, Corvallis, OR
2 Scripps Institution of Oceanography, University of California San Diego, La Jolla, CA
3Kachemak Bay Research Reserve, 95 Sterling Highway, Suite 2, Homer, AK
4Applied Physics Laboratory, University of Washington, Seattle, WA


This study focuses on upper ocean budgets of heat and freshwater, which yield estimates of net surface heat flux and rainfall minus evaporation. The budgets are based on a 19-day ship survey conducted as part of EPIC2001 in September 2001. Underway measurements included temperature and salinity sections from an undulating platform, SeaSoar, and horizontal currents from an acoustic Doppler current profiler along a 146 x 146-km survey pattern centered near 10oN, 95oW in the eastern Pacific warm pool. Additional measurements from a second ship at the center of the survey pattern included radar backscatter from rainfall, air-sea fluxes, and vertical profiles of temperature, salinity, microstructure, and horizontal velocity. Satellite measurements of surface height, temperature and rainfall were also analyzed. The heat budget of 20 and 25-m surface layers indicated that storage, advection, turbulent transport, and penetrative solar radiation were all significant components of the heat budget with a net surface cooling of 41 W m-2 estimated as a residual, which agreed with atmospheric measurements (30 W m-2). The precipitation rate from the freshwater budget was 29 mm d-1, which was in excellent agreement with insitu measurements on both ships and in good agreement with satellite estimates for the same period. Lateral transports of heat and salt were influenced by an anticyclonic eddy in the survey area and it is suggested that anticyclonic eddies, which form near the Central American coast, may carry anomalously warm sea surface temperature toward the west and become preferential sites for heavy rainfall.

 

Air-Sea Interaction in the East Pacific Warm Pool: Heavy Rainfall over Warm Eddies

Hemantha W. Wijesekera and Clayton A. Paulson
College of Oceanic Sciences and Atmospheric Sciences, Oregon State University, Corvallis, OR

In situ, radar, and satellite observations of upper ocean structure and precipitation show heavy rainfall over a warm-core ocean eddy. A warm eddy was identified from both sea surface height (SSH) anomalies in Pathfinder satellite data, sea surface temperature (SST) anomalies in TMI satellite data, and in-situ temperature, salinity and acoustic Doppler current profiler measurements made during EPIC2001. Satellite data showed a ~10 cm positive SSH anomaly moving south-westward from the Central American coast to the EPIC observational site. The diameter of the eddy was about 200 km, and it moved westward at a speed of about 13 cm/s prior to the EPIC survey period. The relative vorticity, estimated from ADCP velocity, was about –0.27f, which is comparable to vorticity estimates based on SSH anomalies. The eddy had a warm SST anomaly in excess of 0.5oC for August and 1.5oC for September. SST anomalies were computed by subtracting 6-year monthly-averaged means from monthly-averaged SST for 2001. Relatively high rainfall along the path of the eddy was observed: the maximum in rainfall associated with the eddy exceeded 30 mm/day during August, and 40 mm/day during September, when both SST and SSH anomalies were largest. A two-day deep-convective event produced 0.2 to 0.3 m of rainfall, which accounted for most of the rainfall observed during the 20-day EPIC ship observations. There was excellent correlation between the location of high rainfall and the warmest SST prior to deep convection.
In addition to the observed heavy rainfall over the warm eddy observed during the EPIC experiment, other sites of heavy rainfall near the Central American coast were identified using satellite SSH, SST, and TMI rainfall. Satellite data indicate higher precipitation over warm anticyclonic eddies in comparison to other areas in the warm pool during months of August and September. Six-year monthly averaged rainfall for the month of September is a factor of two larger over anticyclonic warm eddies compared to that over cyclonic eddies. It is likely that warm eddies, which form near the Central American coast, carry anomalously warm sea surface temperature toward the west and become preferential sites for heavy rainfall.

 

TEMPERATURE FRONT IN THE EASTERN EQUATORIAL PACIFIC NEAR 95W

Hemantha W. Wijesekera1, Clayton A. Paulson1, Daniel L. Rudnick2, Scott Pegau3, and Stephen D. Pierce1

1 College of Oceanic and Atmospheric Sciences, Oregon State University, Corvallis, OR
2 Scripps Institution of Oceanography, University of California San Diego, La Jolla, CA
3Kachemak Bay Research Reserve, 95 Sterling Highway, Suite 2, Homer, AK

This study describes finescale to mesoscale variability across the north side of the equatorial front near 95W. The description was based on two sections of temperature and salinity from SeaSoar, horizontal velocity from acoustic Doppler current profiler, and meteorological measurements from shipboard sensors as a part of EPIC2001. A sharp density front, observed near 0.4-0.5S on 5-6 October 2001 was located approximately 0.75 deg southward from its long-term position. Temperature and salinity changes were about 1.7 oC and 0.2 psu within a 0.7 km distance. Zonal and meridional velocity changes across the front were about 0.3 m s-1 over 3 km. Horizontal and vertical T-S properties agreed well at the frontal zone indicating a subduction of converging water-masses along outcropping isopycnals. Wave-like, mesoscale-fluctuations with wavelengths of about 25 km were found in current profiler records. Internal-wave motions were found across the front and the equatorial shear zone. Wavelengths of these motions were about 0.5 to 1 km. Indirect-estimates of turbulent heat flux indicated that mixing was intense across the equatorial-frontal zone. It is hypothesized that the frontal convergence may be explained by the retention of nonlinear terms in the vorticity equation. The additional terms contain the meridional gradient and curvature of the zonal velocity, which may be dominant terms. A band of observed positive wind stress curl north of the equator provides forcing that is consistent with southward transport north of the front. The combination of this southward transport and northward Ekman transport south of the front is consistent with convergence needed to maintain a sharp front.

 

The Turbulent Near Surface Momentum Balance in the Eastern Tropical Pacific

Yoo Yin Kim, Janet Sprintall, Peter Niiler and Sean Kennan*
The Scripps Institution of Oceanography
*Nova Southeastern University Oceanographic Center

Shipboard ADCP and drifter observations of horizontal velocity, SeaSoar measurements of temperature and salinity, and satellite sea level and wind data are used to compute the June 30-July 15, 1991 time-space mean vertically integrated momentum balance at 95ºW. The KPP (Rev.Geophys.32,4:363:1994) prescribed Richardson Number is used to determine the distance, h, from the surface at which the turbulent stress and its gradient vanish. A consistent vertical average momentum balance to depth h is achieved from a time mean of north south sections’ data on a 0.5º latitude resolution. These data underscore the importance of the southerly wind driven eastward transport of mass in the turbulent layer of the eastern tropical Pacific in the June and July before the North Pacific Countercurrent becomes fully established by the wind stress curl.

 

The Need for Eddy Heat (Freshwater) Flux Divergences in the Stratus Region

Keir Colbo

Analyzing the upper ocean heat and salt budgets for the Stratus region off of South America shows that the budgets cannot balance without a significant eddy flux divergence term. Surface insolation (38 W/m2) is partially balanced by the gyre-scale advection of water equatorward (-15 W/m2). The advection of cool water from the upwelled region to the deep ocean by Ekman transport is only effective within a few degrees of the coast. Beyond this Ekman transport is along SST contours. Vertical diffusion is relatively unimportant, and so the remaining two terms (Ekman pumping and eddy flux divergence) must balance the equation. Ekman transport is convergent over the mooring, and so must represent a heat source. The specifics of which are dependent on the vertical gradients of temperature and Ekman pumping velocity. Consequently eddy flux divergence must be negative and large (> -25 W/m2). Some corroboration for this eddy flux divergence can be found in the mooring and other data sets.

 

Turbulent Entrainment Fluxes in the East Pacific Warm Pool

John B. Mickett and Michael C. Gregg
Applied Physics Laboratory, University of Washington, Seattle, WA

Microstructure and ADCP profiles collected during 19 days in September 2001 at the EPIC intertropical convergence zone (ITCZ) station at 10o N, 95o W are used to investigate the variability and dynamical dependencies of upper-ocean vertical turbulent fluxes of buoyancy (Jb), heat (Jq) and salt (Js). This work is motivated by Wijesekera et. al. [2005] who found the upper-ocean turbulent vertical fluxes to be important in the week-to-month upper-ocean balances of heat and freshwater (salt) in the east Pacific warm pool, as well as by the fact that upper-ocean models are typically sensitive to the parameterizations of the vertical turbulent fluxes.

Time-average values of turbulent fluxes at the top of the seasonal thermocline, or the entrainment depth, he, were {Jb}=-1x 10-8 W kg-1, {Jq}=-8 W m-2 and {Js}=1.5 x10-6 m s-1 psu (or a freshwater flux of -4 mm day-1) (here negative fluxes are downward). Some of the largest turbulent vertical fluxes with Jb < 10 x {Jb},  Jq <7 x{Jq}, and Js >5 x {Js}, were observed when the surface boundary layer (SBL) penetrated the top of the seasonal thermocline, which occurred primarily during two storm events on September 23-25 and September 29-October 1.

Both strong wind forcing (|t|>0.08 Pa) and large, destabilizing surface buoyancy fluxes (Bf > 1 * 10-7 W kg-1) worked together to deepen the SBL to he. The positive surface buoyancy fluxes were due predominantly to the combined effects of large latent heat losses and greatly reduced insolation during storms with heavy cloud cover. More than 1/3 of the total entrainment buoyancy flux Eb= Jb (z = - he) over the 19 days occurred when |t|>0.10 Pa and Bf > 1 * 10-7 W kg-1, which accounted for only about 6% of the time-series. 

The depth of the SBL was reduced (and entrainment often suppressed) by large stabilizing (positive) surface buoyancy fluxes associated with heavy rainfall (> 1 m day-1) and strong daytime insolation (Rs<-1000 W m2). However, the sensitivity of the SBL depth, h, to stabilizing surface buoyancy fluxes was dependent on the strength of the wind forcing. These observations qualitatively agree with boundary layer scaling [Koracin and Berkowicz, 1988] which suggests that h will decrease when buoyant suppression (Bf <0) exceeds wind-driven turbulent production (u*3/[k L]), where L is the Monin-Obukhov length. However, the observations do not fit the theoretical requirement that h<|L| when Bf <0.

A close correlation (r2 È 0.4 at 99% significance) of the entrainment buoyancy flux, Eb, with 8-m shear-squared, Sh28 (and the gradient Richardson number since N2 is nearly constant at he), suggests that the turbulent fluxes are at least partly due to shear-instability. This finding supports parameterizations of turbulent entrainment based on some form of a Richardson number.

Over 60% of the measured shear variance at he is due to near-inertial and low-frequency (w <f ) motions, with the near-inertial and sub-inertial shears interacting to result in a modulation of Sh28 at near-inertial frequencies. This modulation of Sh28 appears to be related to three entrainment events during the first half of the time series that are roughly spaced at an inertial period (69 hours). The connection of Sh28 to the strong near-inertial wave field is supported by contours of Sh28 roughly parallel to theoretical phase-lines of a monochromatic, downward propagating, near-inertial wave.

These observations suggest that resolving or parameterizing both the evolution of near-inertial waves and the sub-inertial background flow may be important to modeling the entrainment fluxes. However, since destabilizing surface buoyancy fluxes working with wind stress generated the largest entrainment fluxes, a focus on improving these bulk parameterizations and the resolution and accuracy of their inputs (winds, rain, solar radiation) should yield the greatest profits.

 

Oceanic Response to Atmospheric Forcing Structure in Eastern Pacific Ocean


Lynn K. Shay, Rosenstiel School of Marine and Atmospheric Science, University of Miami, Miami, FL, nshay@rsmas.miami.edu.


During EPIC2001, oceanic current, temperature and salinity profiles were acquired by deploying expendable profilers from research aircraft flights above the warm pool and along the 95W transect. Analyses of observations support smaller-scale regional variability (i.e. mesoscale) such as wind-stress forced Costa Rica Dome (CRD) that may have extended as far as 110W based on moored TAO measurements. An anticyclonic warm eddy was detected in the warm pool that propagated west to southwest at a speed of 14 cm/s, consistent with Rossby wave. As the eddy propagated in this general direction, it had a pronounced impact on the depth of CRD and on oceanic mixed layer processes. This eddy may be the response to gap winds or formed through current instabilities between the North Equatorial Current and the North Equatorial Counter Current. The key question is how well resolved are these forcing mechanisms in annual, seasonal and monthly mean wind climatologies.

Compared to oceanic climatologies, in situ data exhibits more structure in the upper 100 m where observed buoyancy frequencies (N) are 4 to 6 cph higher across the oceanic mixed layer base. These variations have implications for the OML budgets and SSTs through shear-induced mixing processes in oceanic and coupled models for weak and strong winds (i.e. Juliette). Regardless of the mixing parameterization, if models relax back to a climatology, the T/S structure will not support realistic density and buoyancy structure (or eddies) which will lead to poor predictions of SSTs and OML responses that feedback to the atmosphere during strong wind events such as gap winds and hurricanes.

 

Session 6 - Atmospheric and coupled model experiments

Marine Stratus and Its Relationship to Regional and Large-Scale Circulations

Pingping Xie1), Wanqui Wang1), Wayne Higgins1), and P.A. Arkin2)

1)             NOAA Climate Prediction Center
2)             ESSIC, Univ. of Maryland

A preliminary investigation has been conducted to examine the seasonal variations of precipitation and associated oceanic condition and atmospheric circulations in the NCEP GFS AMIP runs and the CFS Free Runs.  In the GFS AMIP runs, the GFS model is forced by observed oceanic condition, while the same GFS model is fully coupled with an OGCM in the CFS free runs.  Comparisons between the GFS AMIP runs and the CFS Free Runs are able to provide us with insights into how imperfect definition of oceanic condition may influence the CFS model forecasts. Our initial results showed the following:

1)    While large-scale precipitation patterns are reproduced reasonably well over the target regions in both the GFS and CFS, differences exist between the GFS and CFS in the magnitude of precipitation and in the latitudinal position of the ITCZ over both the Atlantic and the east Pacific;

2)     The latitudinal displacement of the ITCZ in the CFS free runs is closely related to the warm SST bias in the SE Atlantic and east Pacific dry zones;

3)    The warm SST bias is primarily a result of the lack of clouds over the region in the CFS free runs (and most other models as well); and

4)    The clouds over the region are mostly stratus with relatively low cloud top and a diurnal cycle with maximum cloudiness observed in early morning.

It is clear from the above results that an improved understanding of the marine stratus clouds is essential to improved simulations, and ultimately prediction of climate.

 

Results from a regional coupled ocean-atmosphere model of the Eastern Pacific.

by Richard Justin Small, Simon deSzoeke, Shang-Ping Xie, Yuqing Wang and Haiming Xu.
International Pacific Research Center, School of Ocean and Earth Science and Technology, University of Hawaii, Honolulu.

We present a 5-year simulation of the coupled ocean-atmosphere-land surface system in the eastern tropical Pacific by the International
Pacific Research Center (IPRC) Regional coupled Ocean-Atmosphere Model (IROAM), a regional atmospheric model coupled to a Pacific basin ocean model. The IROAM simulates the coupled dynamics of the cold tongue, the cross-equatorial flow, and the ITCZ. We focus the analysis on the cross-equatorial flow during the EPIC2001 period. Results from the coupled model will be compared with EPIC2001 observations and previous results from atmosphere-only models.