Acknowledgements: NOAA/OGP, NOAA/OAR

OUTLINE:

* TAO flux moorings
* PACS/EPIC -- Understanding the coupling between the cold tongue & ITCZ
* 95W Enhanced TAO array
* June 2000 observations of the CT/ITCZ system
* Seasonal and intraseasonal variability
* Summary

http://www.pmel.noaa.gov/tao/epic/

The title of this presentation is "Surface heat fluxes from the TAO Enhanced Array along 95W".

The outline of the talk is as follows: I'll begin by describing the TAO flux mooring, and our use of it in the PACS/EPIC experiment.

I'll then show observations and surface fluxes from last June, as the system transitioned from warm season to cool season;
and temporal variability at two key sites -- one in the stratus deck region at 8S, 95W, and some time series to illustrate the seasonal and intraseasonal variability.

I'll end with a brief summary.

This work is funded by NOAA Office of Global Programs and NOAA Office of Oceanic and Atmospheric Research and is part of the Pan American Climate Studies EPIC program. EPIC stands for Eastern Pacific Investigation of Climate Studies.

My Co-PI for this project is Mike McPhaden, director of the TAO project. Most of the figures I'll be showing can be found on the TAO/EPIC homepage, listed here --
http://www.pmel.noaa.gov/tao/epic

Figure 2. TAO homepage (1min/3 min)

Data I'll be discussing come from the enhanced Tropical Atmosphere Ocean (TAO) ATLAS buoys along 95W.

This is the homepage for the TAO project. There are nearly 70 TAO buoys arranged along 10 lines in the tropical Pacific, with the primary mission of monitoring ENSO. Standard measurement on TAO buoys include wind speed and direction, air temperature, relative humidity, SST and subsurface temperature down to 500m. All measurements on the Next Generation ATLAS have at least 10 minute sampling; some measurements on the older ATLAS have 60 minute sampling. Daily averages are computed and telemetered to the laboratory via Service Argos.

To make the ATLAS buoy into a "flux buoy" we have enhanced it with shortwave and longwave radiation sensors, a rain gauge, a barometric pressure sensor, 2 extra temperature sensors in the mixed layer, conductivity sensors to measure salinity at 7 depths between 1 m and 120 m, and 1-2 current meters.

Figure 4. Q0 definition (1 min/4.5 min)

Q0 = (1 - a)Qsw - Qlw_net - Qlat - Qsen

a = albedo (generally taken to be 0.06, Payne (1972))
Qsw = measured incoming shortwave radiation
Qlw_net = epsilon (sigma * SST ^4 - measured incoming LWR)
or use Clark et al. (1974) bulk formula f(SST, Tair, q, cloud cover index)
Qlat = COARE bulk latent heat flux = f(SST(Qsw,Qlw,...), rh, Tair, wind speed)
Qsen = COARE bulk sensible heat flux = f(SST(Qsw,Qlw,...), rh, Tair, wind speed)


Latent and sensible heat fluxes are estimated from a bulk algorithm using high resolution (in our case 10 minute) measurements of wind speed, air temperature, SST, relative humidity, and barometric pressure. Additionally, the COARE algorithm includes a cool skin correction and an adaption of the PWP 1-d model to extrapolate the bulk (1 m depth SST) to the surface. Thus when the warm layer correction is used, the COARE algorithm also requires downwelling shortwave and longwave radiation.

Figure 5. Table of errors (2 min/6.5 min)

This table shows the standard measurement errors for the sensors. Using the COARE flux algorithm we can compute the corresponding rms error by sequentially applying a +/- error value to the measured value.

Notice that this table has mostly near-zeros. The largest errors are caused by wind speed -- a 0.3 m/s error can cause a 4 W/m2 error in latent heat flux, air temperature -- and relative humidity. Rotronics claims the error in relative humidity is 2% which would cause a 10 W/m2 error in latent heat flux. Based on an 1994 analysis of pre and post calibrations we found that the drift was often more like 4% -- which would lead to a 20 W/m2 error in latent heat flux. However, since then, we have acquired a Thunder calibration chamber that has automated the in-house calibration proceedure and eliminate human error in the calibration process. A reanalysis of the pre vs post calibration drift is currently underway.

Of course, these errors do not include the error in the flux by using an inappropriate bulk algorithm.

One objective I have in attending this workshop is to make sure that I have the state of the art algorithm for computing bulk latent and sensible heat fluxes. The fluxes I will be showing were computed using the Fairall et al. COARE bulk algorithm. This algorithm was developed for the convective western Pacific warm pool region. However I will be using it in the eastern tropical Pacific where the atmospheric boundary layer is relatively stable.

Figure 6. TAO flux bibliography (1 min/7.5 min)

There is a long history of using moored measurements to estimate the surface heat, moisture, and momentum fluxes. This is a partial list of some fo the recent publications which specifically use the TAO buoy measurements to estimate surface fluxes. This list is likely to grow rapidly, as PACS/EPIC, PIRATA, and other TAO flux analyses proceed.

This figure shows the cover page of the "Science and Implementation Plan for EPIC:   An Eastern Pacific Investigation of Climate Processes in the Coupled Ocean-Atmosphere System".

The bottom figure shows SST field, with cool (blue color) water off South America and along the equator, and warm waters (represented by reds) north of the equator. As shown in the top cartoon, the meridional structure of the SST field is reflected in the cloud structure: The sratus deck extends from the cool waters off South America to the convective region of the cold tongue / ITCZ complex. At and north of the ITCZ, in the "breeding grounds" of east Pacific tropical storms, SSTs are extremely warm.

Understanding the ocean-atmosphere coupling responsible for the structure and evolution of the CT/ITZC complex is a prime objective of the EPIC program.

EPIC fieldwork includes a short-term intensive process studies (EPIC2001) embedded within longer-term larger-scale enhanced monitoring. The EPIC2001 process study will take place next September and will involve aircraft, 2 ships, aerosonds, etc... Enhanced monitoring, such as what I will be showing, started a little over a year ago.

Figure 8.  Area of work (1 min/10 min)

This is the homepage for the TAO/EPIC project.

As part of the enhanced monitoring for EPIC, the TAO array has been enhanced with 3 extra buoys which extend the line northward through the monsoonal trough, and add resolution on the northern edge of the cold tongue where the meridional gradients are large; and extra sensors to convert each of the 10 buoys into a flux buoy.

Small squares are the TAO moorings.
Big Squares are TAO enhanced flux moorings.
Big Diamonds are new TAO enhanced flux moorings -- at 3.5N, 10N, and 12N, 95W
The Star is Bob Weller's IMET flux mooring in the stratus region.

The enhancements began in November 1999. Final recoveries will occur in the Fall 2003, thus providing 3-4 years of data.

This figure shows the 95W mean cross-section for June 2000. Essentially, this is the observed realization of the cartoon on the cover of the science and implementation document that I showed at the start.

Top panel shows the barometric pressure field,
next panels shows shortwave and longwave radiation.
Then -- rain. Highest accumulations are at 8N.
Below that -- winds and some of the currents (in blue).
-- air temperature, SST and rh.
-- Qlat, Qsen, and net surface heat flux Q0.
The bottom 2 panels show upper ocean salinity and temperature.

During this month, the ITCZ was located between 5-10N. Maximum rainfall occurred at 8N. Curiously, though, the freshest waters are found south of here at 3.5N. Apparently, upwelling associated with the Costa Rican Dome causes the waters there to be saltier than might be expected based on rainfall.

Now lets look at the frontal region between 0 and 2N. At the equator, SST is cooler than Tair and skies are clear, so that radiative forcing for vertical turbulent mixing in the boundary layer almost disappears at the equator. Since there is also almost no surface sensible heat flux there, one can infer that there is only shear-driven atmospheric turbulence, and that this is probably only churning up the near-surface air, not the entire PBL. Because 'ventilation' is so weak, the equatorial rh is very high and surface winds are weak.

However, just north, at 2N, SST increases and is warmer than Tair. As a consequence of the destabilized air column, ventilation from the dry air aloft causes surface specific humidity to be slightly lower, and winds to increase.

Clearly there is complicated meridional structure to the ocean-atmosphere system in the east Pacific. Now let's look at some of the temporal variability.

This figure shows the observables from 8S 95W in the stratus-deck region for the period Dec1999 - May 2001 (present).

Notice the large annual cycle & steady winds. In March there appears to be a SPZC with a reduction in insolation, increased rain, a convergence in the winds, and fresh anomalies forming in a deepening mixed layer.

Latent, sensible heat fluxes were computed from high resolution data and therefore extend only until November 2000, when the data was recovered. Note that a positive latent and sensible heat flux represents a heat loss by the ocean, while a positive net surface heat flux (right axis) represents a net surface heat gained by the ocean.

Warm Season (Jan-Jun):

high Qsw (with big drops during Feb-Mar when have SPCZ),
low Qlat ~ 65? W/m2,
Qsen ~ 0 (hmm, would have expected more so during cool season).
High Q0~150 W/m2 with drops during SPCZ.

Cool Season (Jul-Dec):
lowest Qsw during SPCZ (100 W/m2), next lowest when SST is cooling (200 W/m2), medium (250 W/m2) during period with very low SST.
higher Qlat~100 W/m2 (why? Can't tell if ws is higher from ts plot).
more variable Qsen -- hmm would have thought ~ 0 (stable atmosphere) when have stratus during cool SST period.

Now for completely different temporal variability, let's look at the 10N, 95W data.

All variables are more variable.

We see the ITCZ at and/or north of 10N from Jun 2000 to Nov 2000

We see big variability in thermocline, probably due to variability in the Costa Rica Dome associated with the Papagayo winds.

Qsw lowest (~125W/m2) May Jun 2000
Qlat lowest (10W/m2?) in Dec 99 and (~50 W/m2 or less) in July-Aug 2000
Qsen often near 0 always except in May-Jun 2000 & Sep (when ITCZ overhead)
Q0 changes sign rapidly (AT WHAT TIMESCALE 2-5 days??) during May-Jun 2000

TAO flux moorings can act as flux reference sites. Cross-reference studies with other flux moorings (TAO, IMET, TRITON) are currently underway.

Figure 13. WHOI I/C (1 min/18 min)

This cross-reference study took place onshore at WHOI. The idea was to try to have conditions as similar as possible and have the experiment controlled so that we could minimize natural variability and understand differences in terms of instrumental, sampling, processing, and calibration differences.