WORLD OCEAN CIRCULATION EXPERIMENT
and
TROPICAL OCEAN AND GLOBAL ATMOSPHERE
WOCE/TOGA SURFACE VELOCITY PROGRAMME PLANNING COMMITTEE
Report of the Fifth Meeting
SVP-5
Hamilton, Bermuda
6-8 April 1992
WOCE Report No. 87/92
September 1992
WOCE International Project Office
Wormley
TABLE of CONTENTS
3. REPORTS OF OPERATIONAL CENTRES
4. UPDATES OF NATIONAL AND INTERNATIONAL PROGRAMMES
6. SST OBSERVATIONS FROM DRIFTERS
7. SEA LEVEL PRESSURE OBSERVATIONS FROM DRIFTERS
Annex 1. Agenda
Annex 2. Participants
Annex 3. Air Deployment Instructions
SVP has its scientific roots in the early 1970s when arrays of FGGE drifters were released for studies of mixed layer circulation in NORPAX and EPOCS in the Pacific Ocean. Since then, several statements of scientific goals have appeared. In the November 1991 review of SVP, by WOCE/SSG, these goals were revised and expanded and are presented below:
2.1 Describe mixed-layer velocity on a global basis (500 km x 500 km resolution) resulting in new seasonal surface circulation charts. For Core Project 3 this resolution is to be enhanced to 250 km x 250 km.
2.2 Describe eddy statistics and single particle diffusivities.
More recently the operational meteorological agencies, TOGA, WOCE and the Atlantic Climate Change Program (ACCP) have recognized the importance of surface drifter temperature data for the calibration of satellite SST and sea surface pressure data, especially in the southern hemisphere, for assimilation into numerical weather forecast models. This has led to the additional objective to:
2.3 Provide operational data sets for SST and sea-level pressure.
Drifter measurements of the surface current encompasses a larger group of science (e.g. TOGA) and practical issues (e.g. fisheries agencies) than are of interest in WOCE. The following objectives have therefore been introduced in support of acquiring resources from national agencies and are listed in the Reports of the WOCE/TOGA SVP Planning Meetings.
2.4 Provide a global climate model verification data set:
2.5 Describe the global wind-driven flow and its effects in:
2.6 Provide higher resolution coverage of the:
Common to all elements of the Global SVP is that "WOCE-quality" drifters are used and that the data on ocean mixed layer velocity and SST is interchangeable among the programmes.
3. REPORTS OF OPERATIONAL CENTRES
SVP operations are organized with a Global Drifter Center (GDC) at Scripps Institution of Oceanography, a Drifter Data Center (DDC) at Atlantic Oceanographic and Meteorological Laboratory in Miami and a data centre at Marine Environmental Data Service (MEDS) in Ottawa, Canada. There is a continuing relationship with Service ARGOS and an important relationship with the Drifting Buoy Cooperation Panel (DBCP).
Table 1. SVP Pacific Ocean Commitments and Deployments for 1991
|
|
Array |
|
No. of SVP |
Intl. |
No. of |
|
||||||||||||
|
Calendar Year |
93 |
94 |
95 |
93 |
94 |
95 |
96 |
97 |
||||||||||
|
Equatorial |
15°N-15°S |
|
|
France |
20 |
|
|
|
|
|
||||||||
|
|
40°N-15°N |
|
|
Canada |
16 |
16 |
|
|
|
|
|
|
||||||
|
|
60°N-20°N |
|
|
Brazil |
5 |
|
|
|
|
|
|
|
||||||
|
|
35°S-60°S |
|
|
|
|
|
|
|
|
|
|
|||||||
|
|
10°N-35°S |
|
|
|
|
|
|
|
|
|
||||||||
|
TOTALS |
732 |
693 |
155 |
167 |
98 |
507 |
447 |
523 |
493 |
493 |
||||||||
(*) = "No. of SVP Lagrangian Drifters" is based on 400 day half-life and 95% success rate of deployment
(No. = Array size x 0.90 ÷ 0.95)
(**) = "Equivalent SVP drifters" are the number of sampling blocks occupied. Actual number deployed exceeds "equivalent number"
3.1 Global Drifter Center (GDC)
GDC activities in 1991-92 consisted primarily of drifter deployments, technical developments and distribution of technical assistance to SVP members and the oceanographic community at large. Below is a summary of these activities of which technical developments are expanded upon in Section 5. Table 1 contains an analysis of the Pacific Ocean array commitments and deployments.
(i) Deployments
(ii) Technical Developments
(iii) Technical Assistance
3.2.1 The activities in the DDC in the past year have been centred in the expansion of the weekly produced data quality products to include the mid-latitude Pacific and the North Atlantic Oceans. Data are now also routinely received from Australian, Canadian, Japanese and French drifters in the Pacific. Technical problems with some of the Japanese data set are being addressed by DDC.
3.2.2 The central role of the DDC remains a data assembly centre for the purpose of receiving data from Service ARGOS and processing it to a uniform scientific quality standard. A data archive is maintained at DDC for the use of any WOCE/TOGA SVP PI and this data can be accessed remotely. Each of the national programmes in WOCE and TOGA determines who is a SVP PI. Retrieving data from DDC shows it to be very user-friendly and prompt to answer requests, largely due to the excellent attitude and professional demeanour of the DDC staff. On six month intervals, the data is sent to MEDS.
3.2.3 At the request of WOCE/SSG, a meeting was held 6-7 February 1992 at AOML to discuss the problems of data transfer to MEDS. The discussion at the meeting centred on how data would be transferred to MEDS and the information MEDS would require to keep the WOCE data confidential. The WOCE data policy is that SVP data remains confidential to the PIs for a two year period, unless the PI has authorized their earlier release, which is encouraged. TOGA data, however, must be available in real time for the prediction of ENSO, and a data summary is produced in real time at the DDC and published regularly in the Climate Analysis Bulletin of NMC. SST data goes on to GTS and is quickly incorporated into the global SST analyses produced by NMC/NOAA (and other agencies). GDC will produce a METAFILE for each drifter. The first data tape will go to MEDS by 1 July 1992.
3.2.4 There is an interpolated data file of location and SST (four times per day) at DDC which is produced by application of various interpolating functions to the raw, despiked files. This data has been recently reprocessed, using a modified version of an interpolation scheme which uses a fixed number of raw data points on either side of the desired point. The interpolating functions used are based on the tropical Pacific statistics of temporal structure function obtained from following individual drifters. There is a need to develop statistical routines for each latitude band, because the nature of the high frequency motions change with latitude.
3.2.5 At the October 1991 meeting of the ARGOS tariff agreement conference Service ARGOS agreed to charge 1/3 full service rate if a drifting buoy in SVP is transmitting for any eight-hour period in any 24-hour period. DDC will deploy sixty drifters in the tropical Pacific in 1992 with this format because it gives a more accurate interpolated data set in the tropics. For mid-latitude deployments, however, this sampling scheme can cause aliasing of inertial motions. It is recommended that the one-day full transmission format still be used in the mid-latitude deployments. The special scientific sampling schedule must be specifically requested of Service ARGOS in either case. Users of the reduced duty cycle will lose data from the first pass of each group unless they request Class III service.
3.2.6 TOGA International Project office, in cooperation with the Jet Propulsion Laboratory, is producing a CD/ROM of all of TOGA data for the five year period from 1985-1990. DDC will submit to JPL a tape of the quality-controlled raw drifter data files to be included in this CD-ROM. MEDS has already provided this project with all drifting buoy data they have stripped from the GTS files.
3.2.7 There are several drifter data sets that have been obtained with the WOCE/TOGA Lagrangian drifter that DDC has agreed to put into the SVP files. Participants at the meeting have identified the following data sets:
It is understood that acquisition of this data might take several years because of confidentiality.
3.3 Marine Environmental Data Service (MEDS)
3.3.1 On behalf of the IOC and WMO, MEDS has acted as the Responsible Oceanographic Data Center for global drifting buoy data since 1986. All drifter data that appear on the GTS are collected and archived by MEDS in real time. Simple corrections are made to monthly data files and global maps on the location of sensors on the ocean surface are produced and distributed to some 125 scientific users. For the last year, MEDS has acquired experiment and drifter numbers monthly from Service ARGOS to better manage the GTS data. This has enabled MEDS to make the correspondence between the GTS data and the PIs. The full extent of MEDS activities in this area can be obtained from reports published by the IOC/WMO or from the annual report published by MEDS.
3.3.2 Being the Responsible National Oceanographic Data Center for the drifter data collected under the SVP, MEDS requires the following information from the PIs participating in the programme:
(i) Identification of data to be held confidential; confidential archives are WOCE data or special projects data that use WOCE/TOGA quality Lagrangian drifters. These will be identified by GDC in the META files.
(ii) A META file that contains the complete physical dimensions of the drifter and the calculation of its drag-area ratio, the manufacturer name, the ARGOS number and sensor operation principles will be made by the GDC. Most countries that participate in SVP have agreed that SST data (and later barometer data) is not confidential and can be released by ARGOS on the GTS. MEDS has been working with various countries and PIs, together with DBCP, to make this process work smoothly.
3.4 Drifting Buoy Cooperation Panel (DBCP)
The DBCP is jointly supported by the World Meteorological Organization (WMO) and the Intergovernmental Oceanographic Commission (IOC) and is served by a full time Technical Coordinator. One of its roles is to achieve better collaboration between meteorologists and oceanographers and to promote the use of the Global Telecommunication System (GTS) for real time distribution of drifting buoy data. In the past year, DBCP has been involved in several topics germane to SVP interest.
3.4.1 DBCP supports the work being undertaken at the GDC for the development of an air pressure port for the WOCE/TOGA Lagrangian Drifter for obtaining data from data-sparse ocean areas. A booklet has been written in collaboration with GDC and is being widely distributed among meteorological agencies, which explains how agencies can participate and make commitments for collaboration with SVP. The DBCP has enlisted several agencies to participate in the testing the new barometer ports.
3.4.2 DBCP assists in the development with Service ARGOS of a new system for processing GTS data that allows more flexibility in sensor data processing and automatic quality control. Since the new system will be separate from the existing ARGOS processing system, data delivered to PI (e.g., raw data) can be processed differently from data distributed onto the GTS (i.e., processed data in physical units). The first phase of this project will be operational in December 1992.
3.4.3 DBCP defines and implements deferred-time operating guidelines for quality control of drifting buoy data distributed onto the GTS. This is based on the use of an Omnet/ScienceNet bulletin board named "BUOY.QC" and is intended for a rationalization and acceleration of status change or recalibration process of drifter sensors when needed. BUOY.QC is used when meteorological centres discover errors in GTS data during their operational quality control process (e.g., when observed data do not compare well with first guess fields of various forecast global models and is found problematic). If necessary, the PI is contacted and he requests Service ARGOS to implement a change.
3.4.4 DBCP plans to continue direct contacts with PIs of drifting buoy programmes to seek approval for GTS distribution of their data.
3.4.5 DBCP will support drifting buoy activities of its action groups. The European Group for Ocean Stations (EGOS, North Atlantic) and the International Arctic Buoy Programme (IABP) are two action groups of the DBCP.
3.4.6 DBCP will support in the development of an operational buoy programme in the South West Indian Ocean. This development will be closely watched by SVP.
3.4.7 DBCP will support in the development of a Global Climate Observing System and a Global Ocean Observing System.
4. UPDATES OF NATIONAL AND INTERNATIONAL PROGRAMMES
Argentina presently has no funded drifters as part of their WOCE programme. However, requests by the Argentinean Antarctic Institute have been forwarded for ten drifters in 1993-94 to be placed into the confluence of the Drake/Weddell Sea water masses. Several fisheries agencies also see the need for drifter measurements and are especially interested in aiding deployment of WOCE drifters in their operational areas in the South Atlantic. When SVP begins deployments in the Southern Oceans, Argentina is willing to be a major partner in the deployment operations.
The objectives of the Brazilian drifter deployments are to study the circulation in the Brazilian Current in an area South of Sao Paulo. Brazil will build 17 WOCE/TOGA Lagrangian drifters in 1992 and release them in the study area in 1993. Additional funds will be sought for five more units in 1993.
Germany has released 121 drifters in the South Atlantic in 1990-1992. Twenty drifters which were deployed in the tropical Atlantic have drogues at 15 m and the rest have drogues at 100 m. The emphasis has been to place tight clusters across the major Western and Eastern Boundary Current Systems and to follow the tracks for six months to one year. A request has been put forth for continued funding for South Atlantic deployments at about a rate of 30 drifters a year for three years.
SACLANTCEN has deployed 33 WOCE/TOGA Lagrangian drifters into the Greenland-Iceland-Norwegian (GIN) Seas in the period June 1991- March 1992. Forty-seven more drifters will be deployed in 1992 and requests for 20 more in 1993 have been made. These deployments are along lines that intersect major known surface current systems of the GIN Sea. The scientific objectives of the programme closely follow those of SVP.
Netherlands has deployed three WOCE/TOGA Lagrangian drifters in 1990 and four in 1991. These were placed in the North Atlantic current system, northeast of the Gulf Stream Extension. Plans are to request four to five units in each of the next three years, depending upon the economy that can be reached with Service ARGOS in tracking these devices.
Professor Armando Fiuza at the University at Lisboa has been funded for deployment of drifters according to the plan he presented at SVP-2. Therefore, 30 WOCE/TOGA Lagrangian Drifters will be deployed in 1992-1993 between Portugal and the Canary Islands.
Since 1976, the International Ice Patrol (IIP) of the USCG has air deployed FGGE type drifters for comparison with currents computed from hydrographic data. These drifters have SST sensors and windowshade drogues centred at 58 m. Over 110 drifters have been deployed since 1979, with 8-12 drifters deployed each year. IIP's operational area is 40-52°N, 39-57°W, and is large enough so that dense drifter coverage is not possible. Since the FGGE-configured drifters can be prohibitively expensive for this scale of survey, IIP is interested in converting to WOCE/TOGA Lagrangian Drifters beginning late 1992 (ice season) or 1993. They are in the process of selecting a lead unit for Atlantic and Pacific deployments and also are working on a standardized air deployment package and procedure for release from CG HC-130 aircraft.
The SEMAPHORE project will study the mesoscale ocean circulation and ocean-atmosphere interactions South of the Azores. The field study will occur in July to November 1993. Among the numerous instruments deployed to study ocean and atmospheric circulation are 25 BODEGA drifters. These units satisfy WOCE criteria for drag area ratio and their data will be interchangeable with SVP. SEMAPHORE plans to track those drifters for only two months and it is of great interest to SVP that they have a long battery life so SVP can pick up the cost of tracking after SEMAPHORE interests in them have ceased. It was suggested that SUBDUCTION/WOCE consider making an arrangement with SEMAPHORE for the long-term tracking of these BODEGA drifters.
The US has a drifter programme in support of the Atlantic Climate Change Program of NOAA. Sixty drifters are being released in 1992 in the sub-polar and sub-tropical gyres. In late 1992, 40 more will be released. In the SUBDUCTION programme, 25 have been released in a 5° square around the Azores front, and in 1992-93 45 more units will be deployed. US plans are to request continuation of the deployment of an array of 120 drifters for the 1993-1996 period in the Atlantic Ocean.
Canada's WOCE interests continue in seeding the Alaskan gyre North of 40°N and East of 180°W. Both 15 m drogue and 150 m drifters have been deployed since August of 1990. In 1992, twenty drifters are at hand and will be released through the latter part of the year. Canada's programme will start the second three year funding cycle and a commitment to deploy 30-40 drifters a year will continue through 1995.
Interests of France are in the western tropical Pacific, especially in the TOGA/COARE area. Sixteen drifters will be deployed in August-September 1992, and 15 more in November 1992 and February 1993. It is expected that the funding for 15 drifters per year will continue through 1995.
ROC/Taiwan, in cooperation with US WOCE, has built 27 drifters in 1991 and 36 drifters in 1992. Equivalent funding levels are expected in 1993. These are released into the Kuroshio system south and east of Taiwan in the study of the seasonal cycle of circulation around Taiwan and the East China Sea.
The Japanese programme of drifter deployments is centred in the Hydrographic Office in the Department of Transportation. Drifters are released in several long-term studies. The TOGA commitments continue with 20 drifters per year. In the Northwest Pacific, 4 were released in WESPAC, 4 in the Kuroshio Extension, and 6 along the WOCE hydrographic line at 30°N. This level of commitment is expected to continue into 1993 and onward.
The Korean Ocean Research Institute has released 3 drifters into the Sea of Japan in 1991 and will release 5 more in 1992, with the objective of studying the splitting of the Kuroshio system around Honshu Island. Deployments at this level are expected to continue for several years.
The Shirshov Institute in Moscow has supervised the manufacturing of 18 drifter hulls in 1991 with a commitment of up to a total of 50 before the end of 1992. Funds have been allocated for about 30 hulls in 1993. The GDC will place transmitters and batteries in these hulls and they will be released in the Western Pacific subtropical gyre commencing in late 1992.
US WOCE has a commitment to maintain an array of 132 drifters in the Pacific, and US TOGA has a commitment to maintain an array of 130 drifters in the Tropical Pacific. Continuation of funding for this commitment has been requested for 1993 onward. US will continue to fund the Drifter Data Center and the Global Drifter Center.
SVP is entering into an important new phase of cooperation and coordination with related programmes that deploy meteorological buoys. This cooperation is overseen and fostered by the DBCP, US Navy and NDBC. In each of these three programmes, the objectives have been to deploy drifters with barometers and SST sensors and significant commitments exist in all three operational agencies for an indefinite period of time. The sensors meet WOCE/TOGA SVP requirements, however, presently none of the meteorological drifters have drogues as required in SVP drifters. All three agencies have expressed willingness to use drifters that have SVP specifications of water-following ability. If these operational meteorological drifters were to become drogued, a significant new source of ocean circulation data would become available to SVP (e.g., DBCP coordinates about 100 meteorological drifters, US Navy air-deploys about 600 meteorological drifters and NDBC deploys 30-40 Southern Ocean meteorological drifters). Each of these agencies are closely monitoring the development of the SVP barometer port and the success of utilization of low-cost barometers for measuring sea level atmospheric pressure. Plans have been put forward by US/WOCE to add significant numbers of low-cost barometers which will be deployed in the Southern Ocean by 1993. US Navy has begun the acquisition of the WOCE/TOGA Lagrangian drifter and has plans to attach large-area drogues to their air-deployed meteorological buoy by 1994.
Since late 1990, drifters designed with the principles described in the "WOCE/TOGA Lagrangian Drifter Construction Manual" have been released by the GDC in the Pacific. Data on their survivability over the past 450 days is now available. AOML improved their drogue attachment procedures and data on their survivability was also presented. Figure 1 describes both data sets and shows that the AOML drifters lose power faster than the GDC drifters, although their drogues stay on very well. It is now apparent that to optimize the drifter design for survivability the methods used by AOML in drogue retention should be applied to the GDC surface float and electronics.
Figure 1. Survivability of drifters from July 1990 to February 1992. For (a) AOML drifter (upper panel) and (b) GDC drifter (lower panel) the solid line is an ideal transmission life that is a fraction of drifters that have not gone aground or have not been picked up. The actual transmission life (o), determined every 10 days; the drogue life (x); and temperature life (+) are also shown. It is apparent that both drifters have mechanical survivability of 50% at 400 days but GDC drifter will have produced a more robust data set because the 50% half-life has been reached by the AOML drifter already at 150 days and it stays constant after that. (Courtesy of DDC)
With the cooperation of GDC, the US Naval Oceanographic Office has deployed 14 drifters, 13 of which have had normal life expectancy. The technique of air-deployment is to place the drifter into a water-soluble cardboard box and strap it to a pallet which is covered with a 3/8" thick plywood sheet. An inexpensive 12' diameter marker chute is attached to the box with nylon tubing. Apparently, the box dissolves rapidly upon deployment and the drifter slides into the water. Several times, deployment airplanes have circled the drifter, finding the box riding with 1/4 freeboard out of the water with the chute over to one side. Several suggestions were made to acquire direct observations from a ship on how the drifter escapes possible entanglements from the straps on the box. Clearwater Instrument Co. has used a different technique successfully, whereby the chute falls away from the deployment container upon impact with the water. Annex 3 contains detailed deployment instructions used by NAVOCEANCOM.
5.3 Adaptation of Salinity Sensor
A SEACAT conductivity and temperature sensor has been adapted to the WOCE/TOGA Lagrangian Drifter by the GDC. In TOGA-COARE, 72 units will be deployed and several recovered in the 1992-93 period for post-calibration. The details of an at sea test are described in a report available from the GDC (Swenson et al, 1991). In short, the most serious problem appears to be instrument noise, so at least 10 samples must be taken and averaged over a ten minute period of time to reduce the errors of the mean (to 0.002 psu) due to white noise variability.
5.4 Adaptation of the Barometer
An air pressure port for the WOCE/TOGA Lagrangian Drifter has been designed at the GDC. The port is capable of withstanding frequent immersions without flooding the air pathway to the pressure sensor. To avoid Venturi effects caused by airflow over the surface float, it is elevated approximately 20 cm above the top of the float. The total surface area of the port does not exceed 10% of the total frontal area, so that wind and waves will not induce substantial additional slip producing forces on the drifter. The design is based on a port used on surface moored buoys by the UK Met. Service (Painting et al, 1984), which has had extensive field tests in the wind tunnel. Internal baffling is provided against submergence surges and sufficient back up volume of air assures that water does not enter the barometer duct (Figure 2). At sea tests in 1991 (SVP-4) showed that in 35 knots, (3 m swell and 1 m wind-waves), the surface float with the port attached did not submerge more than a meter under the waves. A fully assembled barometer drifter is depicted on Figure 3.
The prototype barometer drifters are now undergoing final assembly and testing procedures at the GDC. Each of these is equipped with an AIR and a SENSYM SMRT barometer. Both sensors were tested at WHOI for stability for over one year period and they met the 1 mb drift standards. GDC will finish building 25 units by July 1992 and these will be deployed at sea by the Met agencies which purchased them according to the following schedule.
The National Data Buoy Center/NOAA (NDBC) will run the GDC barometer drifters through the standard battery of tests in their environmental test chamber as well as through wind tunnel tests. Each Met agency has their own testing and calibration procedure. The UK Met Office and NMC have offered to follow the quality assessment of all these barometers at sea. The first joint evaluation of the at sea data will occur at the October 1992 meeting of the DBCP in Paris.
NDBC is also developing a barometer system based on KAVLICO and SENSYM sensor elements. A microprocessor is used for data conversion and averaging, corrections in temperature, long term drift, etc. Tests in an air pressure chamber were followed by deployments in the Gulf of Mexico where the initial two months of data on both sensors look very promising when compared to the Paroscientific laboratory standard device. These at sea tests are expected to continue for a year.
6. SST OBSERVATIONS FROM DRIFTERS
From the onset of TOGA, a global SST data set (±0.3°C accuracy, 2° Longitude x Latitude spatial resolution, monthly time scale) was deemed to be vital for the diagnostics, prediction and modelling of ENSO. The need for this data set will continue in all future programmes of climate analysis and prediction.
Commencing in 1985, the construction of a global SST field was taken on by the Climate Analysis Center of NMC/NOAA. By 1990, this product was widely used in not only ENSO research, but also as the input for long range weather prediction in operational centres in the US and Europe. A great deal has been learned at NMC on how to construct a global SST field. Initially, the SST field was based on VOS data reported through GTS, and where no data was available, climatology was blended into the analysis. A few years later, AVHRR retrievals of SST were gradually added, replacing climatology. Drifting and moored buoy data (±0.1°C accuracy) were at first used for calibration of the AVHRR retrieval algorithms. However, as the volume of WOCE/TOGA SVP drifter observations of SST grew, independent verification studies could also be made of both the VOS data and the ongoing, "calibrated" AVHRR retrievals. In this vein, WMO in 1991 completed an assessment of the accuracy of the VOS /SST observing capability. (Kent, E.C., Truscott, B.S., Hopkins, J.S. and Taylor, P.K., The Accuracy of Ship's Meteorological Observation - Results of the VSOP-NA, Marine Meteorology and Related Oceanographic Activities 26, 1991, World Meteorological Organization, Geneva).
These recent analyses of the accuracy of the SST observations and the fields that can be constructed from them bear significantly on the SST observing system required for the continued climate prediction research and long range weather prediction operations. The science objective of long-term SST observations is to produce an in situ data set for the construction of a global SST field of the accuracy and scale set forth at the onset of World Climate Research Programme. At the onset of TOGA it was thought that sufficient in situ and satellite observations would be available to do this. In the waning years of TOGA, quite a different picture emerges.
6.2 Results of TOGA SST Mapping
The principal considerations which emerged during TOGA, and which should be taken into account in the design of the post-TOGA SST observing network, are:
6.2.1 In general, the VOS SST retrievals are accurate to about 1.5°C. This accuracy can be improved, but it requires changes to be implemented both in the instrumentation and the methodology from what mariners are now doing. XBT SST retrievals have larger errors, or noise, than the more common VOS SST reports, because the heat capacity of the probe biases the SST measurements as an XBT first enters the water.
6.2.2 The VOS coverage is not sufficient to span the globe. The retrievals, especially from the Pacific equator and the southern hemisphere, are too sparse to be effectively used to construct maps of SST. The historical data are barely sufficient for constructing monthly climatologies. For example, VOS retrievals did not reveal the 1982-83 ENSO event in real time until it was well into its mature phase.
6.2.3 AVHRR retrieval algorithms are susceptible to 1-3°C errors with changes of atmospheric gas and particulate composition, especially aerosols in the atmosphere. These retrieval schemes can be made stable as long as the aerosol content of the air mass remains stable. Recently, Mt. Pinatubo and Mt. Hudson eruptions has proven to be a major disruption to the retrieval schemes and the Sahara and Asian dust-storms continue to play havoc with the Atlantic and Pacific AVHRR retrievals. This because each dust event represents a different kind of aerosol, inserted at a different level in the atmosphere. Figure 4 demonstrates the effect of both eruptions on global SST retrievals (the biases visible on this figure are also partly due to satellite calibration errors).
Figure 4. The Longitude Average of the in situ Observed SST Minus the AVHRR-Determined Night-time SST as a function of Latitude and Time. The Mount Pinatubo aerosol caused 1.6°C error in SST by September 1991. This error was quickly corrected. The satellite errors from November 1991 through February 1992 are partly caused by the aerosols from the eruption of Mt. Hudson and partly caused by a satellite calibration error. These errors were made worse by the Pinatubo correction algorithm. The contour interval is 0.5°C. (Courtesy of R. Reynolds)
6.2.4 The most accurate real time SST data set today comes from drifting and moored buoys. The NMC SST product now relies heavily on buoy data in the construction of the global maps, not only in areas where there are no VOS observations, but also in screening erroneous VOS reports from heavily traveled shipping lanes. NMC considers buoy data essential to producing and verifying the accuracy of the global SST fields to the stated accuracies put forth in the onset of TOGA. NMC would like to have more buoy data than it now receives through GTS.
6.2.5 The most accurate and best resolved in space and time SST products now available use a technique which combines AVHRR data and drifting and moored buoy data. RSMAS has produced SST fields in the tropical Pacific, with the joint use of drifters and AVHRR which can be shown to satisfy the ±0.3°C accuracy requirement. NMC SST field is now produced by using the sparse buoy data to remove large scale biases and trends in the AVHRR retrievals and to screen the VOS observations. Then the smaller scale two dimensional curvatures of the SST field are established by the AVHRR retrievals and an objective blending of all the data is used to produce an SST map.
Figure 5. North-South Spatial Scales (upper panel) and East-West Spatial Scales (lower panel) of the errors associated with Satellite Retrievals of SST. These errors were computed from correlations of weekly satellite fields for a one year period. The contour interval is 100 km. (Courtesy of R. Reynolds)
6.3 Observations Required for Accurate SST Maps
From the TOGA experience on the production of a global SST field, it is apparent that in situ, high quality observations (±0.1°C) must be made on a global basis in order to produce an SST field to the required accuracy and scale for ongoing climate research. NMC estimates that one in situ device on the average in every 5° square is required to "tie down" the AVHRR retrievals on a global basis. This estimate is based on the horizontal coherence scale of the satellite retrieval errors (Figure 5). It is somewhat larger than the scale of inter-annual SST anomalies. Some places require a more dense buoy network, while in others high quality VOS data can be used. About 1000 buoys in the global oceans are required to supplement the VOS data from high density shipping lanes. This scale and scope of SST observations from buoys are identical to the WOCE and TOGA requirement for surface current data, and it is also the scale on which operational weather centres require sea level atmospheric pressure over the remote oceans.
The specific number of drifters required will depend upon several factors:
(1) The moored measurements in place at any particular time and at any particular place in the world,
In the absence of improved VOS measurements and moorings, 900-1000 drifters would be required. VOS improvements will take another 5 years to implement. Also, drifters serve the purpose in the TOGA array to determine the field of surface currents. If a field of surface currents was not needed in TOGA, then the TOGA TAO array would serve the purpose in the tropical Pacific.
6.4 The Role of Drifting Buoys in SST Observations
It is apparent that an in situ SST observing system is required to be maintained in the global ocean which is capable of measuring SST in real time to 0.1°C accuracy and is deployed on a space scale of the errors of remotely sensed (e.g. AVHRR) and VOS-SST retrievals. The objective is to produce SST maps of ±0.3°C accuracy.
The understanding of how 30-1000 day time scale changes of the tropical SST force the evolution of the global atmosphere was one of the basic research tasks of TOGA. Several physical mechanisms underlying this interaction are now well understood statistically, observationally and theoretically. A hierarchy of models which use the current state of the global SST to predict the joint evolution of the global atmospheric ENSO and the Pacific Ocean tropical El Niño 6-12 months in advance now show significant skill.
In the next ten years, coupled ocean-atmosphere climate modelling and prediction efforts will intensify and all of these future studies will require knowledge of global SST to an accuracy which resolves the significant SST anomalies on a global basis. In the western tropical Pacific and the eastern tropical Indian Ocean, the ENSO time scale anomalies are 0.2 - 1.0°C. In the sub-polar North Atlantic, the SST anomalies associated with interdecadal periods, have amplitudes of 1.0 - 2.0°C. Therefore, it is appropriate to constrain the SST product to a 0.3°C accuracy for future climate event studies. Analysis for the signature of global warming, or trends in SST, also requires a verifiable, and well resolved SST field over large parts of the globe. Here, in situ observations play even a more critical role because no one knows where the global warming will first be manifest.
Secondly, operational, extended weather prediction efforts under way now at several national and international meteorological centres require global fields of SST of comparable accuracy with climate studies. Also climate prediction, as has been made in TOGA, will become operational in the post-TOGA period, imposing an operational requirement for global SST field of TOGA quality. More regional, and more highly spatially resolved SST fields, but perhaps less accurate that the climate forecast data, are required for weather prediction. In this latter effort, the less accurate but high density VOS observations, combined with satellite retrievals play a more important role. Both of these operational requirements can significantly benefit from a permanent array in situ SST observing system which meets the climate-quality data requirements.
6.4.2 The in situ SST Observing System
For continued climate research and long-range weather prediction, the accurate, in situ SST observations will come from three sources: (i) improved VOS in selected shipping lanes, (ii) TAO moored buoys in the equatorial Pacific and (iii) basin scale arrays of Lagrangian drifters. A global array of 900-1000 drifters are required to provide the accurate SST data where ships do not go. By the end of 1992, an array of 470 drifters (jointly coordinated by SVP and DBCP) will be operational on a continuous basis in the global ocean, so, drifter resources must be increased by about a factor of two to meet the requirements for accurate SST observations. The continuous array of approximately 470 drifters is arrived at by adding all the commitments made in the Pacific and Atlantic by SVP participants. In the next five years, drifter measurements will be coordinated with the needs of TOGA, WOCE and ACCP; additional drifter data contributions will come from various agency programmes, such as Fisheries, Navy, Coast Guards and environmental protection agencies. After 1997, a fully operational network should be in place.
Presently, funding requests have been made to several agencies from SVP participants to expand SVP observations from the present level of 470 drifters to the required level of 900-1000. It is anticipated that this expansion will occur over a five year period from 1993 to 1997. Both research scientists and operational agencies will play an important role in the expanded drifter SST observing capability.
7. SEA LEVEL PRESSURE OBSERVATIONS FROM DRIFTERS
7.1 Historical Drifter Observations of Sea Level Pressure
WCRP has considered global sea-level pressure observations to be important for the analysis of ENSO time scale phenomena and has therefore encouraged enhancement of these observations, especially from the southern hemisphere. Several countries have contributed to maintaining an array of FGGE-type drifting buoys, without drogues, in the Southern Ocean from 1984 onward. The data from these buoys are used for both southern hemisphere weather prediction as well as ENSO forecasting (Neville Smith, The Impact of In Situ Marine Observations in Surface Flux Estimates derived from Operational NWP, Discussion paper for the Surface Layer Scientific Panel Meeting, ECMWF, 25 October 1991).
The largest number of drifting buoy observations was done in 1979 during FGGE when 180 buoys were reporting usable data simultaneously. The impact of FGGE buoy data on southern hemisphere analyses (Guymer and Le Marshall, 1981) shows that there was a significantly enhanced description of the number and intensity of low pressure systems around Antarctica. With FGGE buoy data, these circumpolar perturbations appeared more vigorous than the climatological variability had been until then. Van Loon (1980) describes the problems of computing sensible eddy heat transports in the southern hemisphere with and without FGGE buoy data. The enhanced low pressure systems mapped by the FGGE buoys gave larger amplitudes to these fluxes than had been obtained with climatological, low resolution assimilated data.
From the point of view of modelling ocean circulation, it is also apparent that the analyzed synoptic winds without adequate sea level pressure data are weaker than in reality. Smith (1991, see above) has shown with the 1991 version of ECMWF model that significant differences of synoptic surface wind stresses result when the surface observations in the southern hemisphere are eliminated. Over 80% of the surface observations in the southern hemisphere today come from the drifting buoy network. Similar results pertain to estimates of synoptic latent heat flux.
It is not yet known how measurements of pressure would affect the 500 mb height or the number of storms in the Southern Hemisphere with today's increased addition of satellite data and especially attempts to bring scatterometer data into low level assimilations. However, if all the marine surface observations are excluded (Smith) with the present utilization of scatterometer data, the impact on derivation of ECMWF surface fluxes seems to be significant.
It is thus clear that when significant number of drifting buoy sea level pressure observations are added, or when existing ones are removed from the real time data that is assimilated into atmosphere circulation models, large changes of the computed interaction between the ocean and the atmosphere result. From 1993 onward, SVP has plans to expand the drifter network into the Southern Oceans. To reach the scientific goals outlined in Section 1, the network of pressure observations from these drifters must be brought to at least the FGGE standard. WOCE modelling of ocean circulation and SVP interpretation of the causes for surface currents require high quality surface winds from which stresses can be computed.
Figure 6. Difference between ECMWF and NMC Sea Level Pressure Analyses Averaged over June through August 1986. Upper panel is RMS of difference, lower panel is maximum value. (Courtesy of Oregon State University)
7.2 Need for Sea Level Pressure Observations in Satellite Altimetry
In 1992, the TOPEX/POSEIDON mission will be launched. This satellite will carry a radar altimeter and two of the most important algorithms (among the 30 required to render backscatter to sea level height) depend upon an accurate data on sea level pressure. The first is a dry tropospheric range correction due to atmospheric gases, primarily oxygen. This refractive index correction is equal to 0.228 x P (P = sea level pressure) whence a four millibar error in P will lead to one centimetre error in sea level height.
Secondly, an inverse barometer correction that is due to the static effect of atmospheric pressure loading on the sea surface must be made. In general, the relationship between P and inverse barometer response is time and space dependent. However, for time scales of two days to two weeks, it is appropriate to correct the sea level by -1.01 cm/mbar.
Presently, the sea level pressure analyses by different meteorological centres, for example, ECMWF and NMC, differ significantly. Figure 6 displays this difference during June through August 1986 where RMS differences exceed 6 mbars in the Southern Ocean.
Satellite altimeters will be flown by various agencies for the foreseeable future. The interpretation of their measurements for large-scale ocean dynamics requires better sea level pressure analyses than is now available from existing operational centres and their models in the Southern Oceans. The easiest and most accurate way to improve this is to improve the quantity of sea level pressure observations.
7.3 SVP Plans for Southern Ocean Barometer Deployments
In 1992, twenty-five SVP low-cost barometer drifters will be deployed in tests around the globe (Section 5.4). After the initial tests, continued plans are to deploy a second test array into the Southern Ocean in 1993, in cooperation with NDBC. The Southern Ocean is defined for our purposes as being between 35°S and 60°S. By 1994, it is anticipated that the present arrays of barometers that are deployed by various national MET agencies and coordinated by DBCP (Section 3.4) will increase, because the low-cost barometer drifter will cost less than half of what MET agencies are now paying for FGGE-type drifters. The number of drifters presently in the Southern Oceans is between 40 and 60: NDBC maintains approximately 40 drifters, Australia 10 and New Zealand 3. By 1995, a full contingent of 180 low-cost barometer drifters, or the FGGE scale array, should be fully deployed in the Southern Ocean. The 180 drifter array by 1995 is partly based on the success of the upcoming tests. However initial prototype tests of SVP drifters fitted with a barometer were promising. Funding for the maintenance of the Southern Oceans array has been requested from NSF and NOAA. The successful implementation of these plans depends upon continued close cooperation between DBCP, SVP and the newly and rapidly developing Global Ocean Observing System and the Global Climate Ocean Observing System.
Field experiments were conducted during 1990 and 1991 to evaluate the drift characteristics of different buoy/drogue combinations by deploying and tracking arrays of buoys. FGGE-hull/window shade, WOCE/holey sock (15 m and 58 m drogue centre), and other minibuoys were evaluated. A meteorological buoy with a window shade drogue was part of each array. During each drift period, one FGGE-hull drifter had an S-4 electro-magnetic current meter above and another below its drogue to measure slip directly. During 1991, the meteorological buoy had S-4 current meters suspended below it at 3 m and 15 m.
Hydrographic casts were conducted daily in the area of the array. The array was evaluated in the strong Labrador current (50-70 cm/s) east of the Grand Banks and the wind-driven-dominated current (20-40 cm/s) of the northeastern Newfoundland Shelf. After two different 7 day drifts (one 100 km and one 200 km) in the Labrador current, array dispersion was small (10 km). The range of slip velocities for the FGGE-hull was 5 to 15 cm/s during periods of 1.5-3.2 waves and 4-10 m/s winds measured (at a 3 m height).
As part of the ONR Eastern Boundary Current Accelerated Research Initiative, we plan to make a sequence of Lagrangian drifter deployments to study the near-surface flow regime and eddy field in the California Current. Satellite-tracked drifters will be used to characterize the mean flow and its variability (including the Reynolds stresses associated with the eddy field) and to investigate the local dynamics of individual eddies through smaller-scale deployments, with the overall scientific objective being to develop a better understanding of the dynamical role of the eddy field with respect to the California Current system, and a coherent array designed to sample an individual eddy. The incoherent array consists of 11 drifters spaced approximately 20-40 km apart along an east-west transect at about 39.5°N. This spacing was chosen based on CTZ results to given independent trajectories from launch. The coherent array consists of 12 drifters to be deployed in a spiralling squares pattern that gives initial separations of 8 to 84 km. We plan to make six deployments of the incoherent array over a one year period starting June 1993 and two deployments of the coherent array during hydrographic surveys in June and August, 1993. Two types of drifters will be used in these deployments. Most (66) will be standard WOCE SVP drifters drogued at 15 m. The rest (24) will be Metocean drifters supporting Saclantcen ocean colour sensors to measure surface downward irradiance and subsurface upwelling irradiance. These drifters will have WOCE SVP drogues at 15 m, and are being developed by Mark Abbott at Oregon State University to study phytoplankton variability in the California Current system.
GLOBEC is a long-term NSF/NOAA funded research programme presently being developed to study basic physical/biological interactions and recruitment dynamics during the early life stages of several key species. The first major GLOBEC field experiment will be conducted in the Georges Bank/western Gulf of Maine region during 1994-98, with intensive measurements to be made in January-July, 1994-1996-1998. The key species in this study are cod and haddock larvae and several zooplankton including Calanus Finmarchicus, and the overall objective of the study is to identify and understand the physical and biological processes which control the distribution and abundance of these key species during their pelagic stages. While intensive planning for the Georges Bank field experiment is just beginning, some scientific questions have arisen that may be best addressed through Lagrangian drifter experiments. These questions include the following: What is the seasonality of the around-bank flow? When does the around-bank gyre become closed (i.e., when does recirculation occur, and over what region of the bank does this happen)? What are the source regions for the water (and zooplankton) found on the Bank? What are the primary mechanisms that govern the loss of water (and key species) from the Bank? For example, do storms tend to drive off the surface water from selected regions of the Bank, thus possibly carrying away a significant fraction of the key species biomass? Since drifters will tend to leave the Georges Bank region relatively quickly after deployment (say within 1 or 2 months) and enter the slope water region, it is hoped that potential GLOBEC drifter experiments can be coordinated with future WOCE SVP and NOAA Atlantic Climate Change drifter experiments so that both programmes benefit from any coastal deployment of drifters in GLOBEC.
Between fall 1989 and end of 1991, about 50 surface drifters equipped with a mini thermistor chain of 20 meters were deployed in the Tropical Pacific (Figure 7). We analyze temperature records which exhibit strong variations of the diurnal cycle. These variations in the western equatorial Pacific are especially strong during occurrence of westerly bursts. We analyze near surface time series in relation with available wind measurements. In the western equatorial Pacific, the amplitude of the 2 metres' diurnal cycle does not exceed 0.8° during light winds. At 20 metres, the amplitude is 0.1 to 0.2°C. A time series during a short westerly wind burst (7 m/sec) in June 1990 shows the absence of diurnal cycle, an intensification of the mixing and a linear decrease of the temperature between 5 and 20 metres - a temperature inversion then exists, probably due to precipitation, which corresponds approximately to a 0.158 su decrease of salinity between 2 m and 20 metres.
Figure 7. Schematic of BODEGA Drifter, fifty of which have been deployed in the Tropical Pacific by SVP France. (Courtesy of Y. du Penhoat)
The TOGA Numerical Experiment Group proposed a comparison exercise on the seasonal cycle in the tropical Pacific ocean. The wind stress forcing is calculated from monthly mean data of Hellerman and Rosenstein (with a reduction factor of 0.75). The heat flux is taken from Oberhuber's atlas with his specified form for the restoring term. Many groups were involved in this exercise and two different parameterizations for vertical turbulent mixing will be discussed within this framework. The first parameterization follows the scheme proposed by Pacanowski and Philander (1981) and the second one is based on a prognostic equation to calculate TKE (turbulent kinetic energy) and a diagnostic closure to estimate the mixing length. This new scheme is associated with a penetrative solar radiation algorithm. The latter physics permits to reduce biases in SST over large areas: the ocean gets colder as mixing is more efficient but the equatorial upwelling is reduced. By increasing the equatorial Ekman depth, the momentum balance is modified in the upper ocean: vertical shear is reduced and subsequently, equatorial divergence and upwelling are reduced. The TKE scheme strongly affects the equatorial circulation but a definite answer on its properties will require a quantitative comparison with observed surface currents over the whole basin.
The Lagrangian and Eulerian descriptions of the flow in a double gyre, eddy resolving numerical simulation are compared in the context of exploring the use of drifter arrays to describe ocean circulation. The parameters that determine the model ocean circulation were chosen such that the mean and eddy kinetic energy levels are comparable to observations in the upper ocean. The number of "buoy-days" used in the numerical experiment is similar to what is expected to be launched in the Atlantic Ocean during WOCE/TOGA/Global Climate Change surface velocity programme. By using a combination of Lagrangian and Eulerian statistics, it is observed that with a large number of particles the mean Eulerian velocities and velocity variances can be estimated well from the Lagrangian trajectories. The estimation of Lagrangian statistics (i.e. dispersion rates with respect to the centre of mass, Taylor diffusivities, etc.) depends significantly on the region in which they are computed. The estimation of the spatial distribution of the diffusivity function from the trajectories of the particles released in the eddy resolving numerical model accurately reproduce the most important large scale characteristics observed in the analysis of drifters and floats in the ocean: anisotropy of the horizontal components of the diffusivity matrix with zonal values usually being larger than meridional diffusivities; and an inhomogeneous diffusivity field, with large values in those regions where the eddy kinetic energy is larger. Central gyre statistics are typically well defined both in terms of the theory and within the drifter densities used. In the western boundary layer Lagrangian statistics are not robust, not because of sample size problems, but due to breakdown of the assumptions behind single particle calculations. Regimes where this occurs have ratios of local advective time scale to the Lagrangian decorrelation time scale greater than one and are therefore typically non-stationary. Comparison of drifter derived representations of diffusion and other schemes with observed tracer fluxes in the model gyre suggest that drifters and floats can provide a major improvement in our knowledge of ocean diffusion, but that further advances in the way we use Lagrangian statistics are needed.
Figure 8. Tracks of Drifters in the Northeast Atlantic from the Netherlands' Buoy Releases in 1991-92. A mark is indicated every 10 days. (Courtesy of L. Otto)
The estimation of the spatial distribution of the diffusivity function from the trajectories of the particles released in the eddy resolving numerical model accurately reproduce the most important large scale characteristics observed in the analysis of drifters and floats in the ocean: anisotropy of the horizontal components of the diffusivity matrix with zonal values usually being larger than meridional diffusivities; and an inhomogeneous diffusivity field, with large values in those regions where the eddy kinetic energy is larger.
Drifters were deployed in the framework of a programme that aims at the study of the current structure and transports of water and properties in the NE Atlantic region (Figure 8). Analysis of the drifter tracks in combination with the hydrographic data during deployment supports the idea of a broad frontal structure with persistent mesoscale eddies and localized countercurrents.
Analysis of dispersion characteristics for separate winter and summer conditions gives indication for stronger dispersion in winter on the larger spatial scales, but statistically the available data do not allow firm conclusions.
*Deep and Upper Transport, Circulation and Hydrography, WOCE Atlantic Research Programme.8.8(a) Drift measurements in the Ekman Layer, W. Krauss, (Institut für Meereskunde, Germany)
A set of buoys has been deployed in the North Sea drogued in 3-13 m, 7-17 m, 12-22 m, 17-27 m, and 22-32 m in order to measure the drift within the Ekman Layer. The experiment lasted from 20 November 1991 to 28 February 1992. A preliminary analysis of the data is presented.
8.8(b) Drift measurements in the North Atlantic, W. Krauss (Institut für Meereskunde, Kiel, Germany)
A large number of drifting buoys drogued in 100 m depth have been deployed in the North Atlantic Ocean from 1981 to 1990. The analysis of this data set, which consists of 35,000 buoy days, is nearly completed.
The main results are:
Figure 9 shows the mean drift vectors for all drifters deployed in the North Atlantic from 1981 to 1990.
Figure 9. Mean Drift Vectors of FGGE-type Drifters Drogued to 100 m depth in North Atlantic. Resolution is 2° x 3° spatial average and ensemble time average from 1981 to 1990. Deployed in the North Atlantic by Professor Krauss and his colleagues. (Courtesy of W. Krauss)
8.9 Drifter Dispersion in a Sea Current Field, Nikolay Maksimenko
The first test of WOCE drifters manufactured in the USSR was fulfilled in October 1990 in the Northeastern Black Sea in collaboration with GDC. Five day drift of six buoys revealed fairly stable along-shore current with weak eddy activity (Figure 10). Being launched within a few hundred metres from each other, they moved in a rather compact cluster. Having passed more than 60 miles, the drifters increased their relative distance to only a few kilometres. Five of them were recovered at the same convergent zone easily seen from the ship because of the ripple absence. The last one had lost its drogue and was moved away from the cluster by the current vertical shear of about 30 cm/s per 15 m depth. Convergent zones seem very important for drifter trajectory analysis and their dynamics requires careful study.
Kinematic modelling of two-dimensional particle diffusion and tracer field stirring revealed the maximum of their intensity in the eddies field evolving in the time scale corresponding to in-eddy circuit time, which is typical for mid-ocean mesoscale currents.
Analysis of the possibility of diffusion coefficient K parameterization showed that it can be fruitful only when the accurate value of K is not important.
Figure 10. Drift Path of Russian and Ukrainian Drifters in the Black Sea during tests of Drifter Hull Performance in October 1990. (Courtesy of N. Maksimenko)
Between January 1990 and June 1991, 64 ARGOS drifters were deployed in the 106-Mile Deepwater Municipal Dumpsite, which is located in the Slope Water region (39°N, 72°W) north of the Gulf Stream. Drifters were deployed at approximately 1-week intervals as part of a large EPA monitoring programme to determine the fate and transport of sewage sludge dumped at the site.
Drifters were manufactured by Clearwater Consultants and configured with 1-m diameter, 7-m length holey-sock drogues. Drogues were tethered to a depth of 10 m to reside in the mixed-layer; temperature sensors were installed in the surface buoy. Drifter positions were obtained via Service ARGOS at an average of 8 times per day. Drifters were programmed to cease transmitting after 120 days. For the total programme, 7,186 days of trajectory data were obtained.
Surface currents in the Slope Sea carried all drifters southwestward, along local isobaths, to the north wall of the Gulf Stream. Once associated with the Gulf Stream flow, all moved rapidly northeastward with the majority being carried into the distant North Atlantic. Approximately 20%, however, became associated with eddies that broke away from the north wall of the Stream, entered the Slope Sea, and recirculated back toward the southwest in agreement with the hypothesized Slope Sea Gyre.
Three aspects of the observed drifter motions are presented: (1) the time-varying structure of the velocity field and the inferred farfield fate of fine particles released at the 106-Mile Site, (2) the Lagrangian statistics of the surface velocity field in the Slope Sea, and (3) the large-scale dispersion calculated from Lagrangian statistics.
Velocity and SST data from Lagrangian drifters deployed in the TOGA Pan-Pacific Surface Current Study between July 1988 and March 1991 are used to estimate the total rate of change SST, (DT/Dt), following the drifters. Independently, the horizontal advection of mean temperature by the mean flow is computed from the maps of mean temperature and velocity: (ud/dx+vd/dy)(T). The difference between these functions is the net eddy temperature flux convergence DEL.(V'T') due to the three dimensional eddy velocity V'. The net balance implies that cold water particles which rise on the equator due to eddies simply move off the equator and sink as cold water without absorbing any heat from the atmosphere or the mean circulation of the upper ocean. Horizontal heat advection in the powerful South Equatorial Current is no larger than the advection near OWS Papa in the middle of the subpolar gyre
A repeated release experiment using SVP-standard surface drifters is being carried out in the subduction region of the northeast Atlantic Ocean. Twenty-five drifters were deployed between July and December 1991 in the vicinity of the Azores front between the islands of the Azores and Madeira. Forty-five additional drifters will be deployed in 1992. The goals of the drifter measurements in this region are to observe the two-year mean surface currents and, if possible, their large-scale (~1000 km) divergence in a region of outcropping density surfaces. The long-term mean currents and divergences will be compared to Ekman transports from moored meteorological measurements. The measurements are one component of cooperative programme called the SUBDUCTION Experiment sponsored by the Office of Naval Research. Principal investigators are Dr Pearn P. Niiler and Dr Jeffrey D. Paduan.
It is hypothesized that ocean fronts play a major role in the surface layer processes that act to modify density and thickness properties of subducting layers in the outcrop region. To investigate this hypothesis, drifter deployments are being concentrated both north and south of the climatological location of the Azores frontal zone (approximately 35°N). Five instruments are being released along both lines in each deployment cycle. Convergence toward the frontal zone will be monitored and compared for the northern and southern drifters as will eddy variability and diffusivities.
Technical observations from the first 25 instruments have provided some useful guidance. It has been shown that:
Preliminary scientific results have shown that:
The field phase of the SUBDUCTION Experiment will continue through 1993. Drifter measurements in the area will be extended beyond that date by the French SEMAPHORE Experiment, which plans to deploy approximately 25 SVP-standard drifters in the same area. All drifters are (or will be) equipped with enough batteries to continue transmissions while outside the specific area of this experiment in order to contribute to the larger-scale SVP array in the North Atlantic. All data are being collected and managed in real time by the SVP Data Assembly Center in Miami. Sea Surface Temperature (SST) data are being disseminated in real time to the GTS for use in worldwide weather prediction schemes. In addition, Don Olson of the University of Miami is using the SST data to ground-truth satellite AVHRR images in the SUBDUCTION region.
8.13 Seasonal Cycle of Surface Circulation in the Eastern Tropical Pacific, Pierre-Marie Poulain
The trajectories of satellite-tracked drifters during the period from 1979 to 1990 are used to study the seasonal variability of the surface circulation in the eastern tropical Pacific. The mean seasonal (annual and semi-annual period) and high-frequency velocity fields are estimated using a multiple linear regression as a function of latitude (Figure 11).
Figure 11. Monthly Mean of Zonal Currents (upper panel) and Meridional Currents (lower panel) averaged over 180°W to 80°W Longitude as a Function of Months and Latitude. Drogued drifter data from 1979 to 1980 was used in compositing velocity estimates in 2-degree latitude bands. (Courtesy of P. Poulain)
The mean zonal currents are westward north of 9°N (NEC), eastward between 9°N and 4°N (NECC, with a maximum speed of 30 cm/s), and westward south of 4°N (SEC) with two maxima near 2°N (35 cm/s) and 5°S (25 cm/s), and one minimum (10 cm/s) near 1°S. The meridional circulation is essentially poleward with a typical speed of 5 cm/s.
In the zonal direction, both annual and semi-annual signals are maximum near the equator where they reach 30 cm/s and 10 cm/s, respectively. The annual component has a second peak around 9°N which is out of phase with the equatorial maximum. The mean and seasonal components can account for more than 50% of the total velocity variance. At the equator, they combine to give maximum eastward (westward) flow in May (September). At 9°N, zonal currents are maximum westward (eastward) in May (October). No significant seasonal variability is apparent for the meridional circulation.
8.14 SOFAR Float Trajectories in the Tropical Atlantic, P.L. Richardson and W.J. Schmitz, Jr. (WHOI)
We have just obtained the first SOFAR float trajectories in the tropical Atlantic. These provide a direct measurement of cross-equatorial flow of water and of the connections between western and boundary currents and equatorial currents. The cross-equatorial flow which is generally northward in the warm upper layer and southward in the colder lower layer is a major part of the large scale conveyor belt of thermohaline circulation that results in a northward heat flux through the Atlantic and which is important for world climate. Because of the lack of direct observations, the pathways, rates of exchange and temporal variability of the thermohaline circulation have remained poorly known.
We investigated the cross-equatorial flow by tracking floats for 21 months in the North Atlantic Deep Water at 1,800 m and 3,300 m and in the Antarctic Intermediate Water at 800 m. Eight floats at 1800 m drifted swiftly southeastward paralleling the continental slope in the deep western boundary current. It was found to be a narrow, 100 km wide jet, flowing with peak speeds of 55 cm/s and average speeds (10 km bins) of 25 cm/s. Its transport estimated by combining the float observations with current meter measurements obtained by C. Colin amounted to 15x10 6 m 3 /s (from 900 m to 2,800 m). Roughly 6x10 6 m 3 /s of this transport recirculated between the jet and the Mid-Atlantic Ridge leaving 9x10 6 m 3 /s to cross the equator. The trajectories showed that at times the current turned eastward and flowed along the equator and at other times the current crossed the equator and continued southward. Thus the cross-equatorial flow appeared to be linked to the direction of flow near the equator which varied interannually. No obvious western boundary current was observed by the 3,300 m floats which suggested a low mean velocity at this depth. Thus the deep western boundary current was split vertically into two cores, one above 3,300 m and one below (the deeper one was not observed by floats).
A 300 km wide western boundary current was observed to be going northwestward at 800 m in the Intermediate Water. Several floats that drifted in this current looped in anticyclonic eddies that translated up the coast. These eddies are inferred to be pieces of the North Brazil retroflection that separate near 7°N and extend at least from the surface to 800 m. Six 800 m floats drifted eastward along the equator between 5°S-6°N at a mean velocity of 11 cm/s. One float reached the Gulf of Guinea near Africa indicating the equatorial current extended at least 35-40° along the equator. Three floats near the equator reversed direction near the end of the tracking implying interannual variations.
One hundred twenty eight mixed-layer drifters (100 of which were equipped with thermistors) were deployed during the spring and summer from 1985-1988 in the California Current System (CCS) as part of various observational programmes (Figure 12). Altogether, these tracks represent almost 25,000 buoy-days of data. We compute the Lagrangian estimates of the velocity statistics from the long-term tracks of these drifters between about 39°N and 21°N in the CCS. The mean description reveals a California Current that flows southwestward to about 30°N that then turns southward between 24°N and 30°N and finally veers southeastward in the southern part of the domain. There is also evidence of recurrent eddy off of Punta Eugenia (28°N). The variance field is remarkably isotropic with considerable structure in strength. Conspicuous features include regions of high coastal variance (36°N and 26°N), a region of strong offshore gradients (24-26°N), a region of small offshore gradient (30-32°N) and a north-south minimum at 24°N. These results provide a strong constraint for numerical models of the CCS.
Figure 12. Drifter Tracks 1985-90 in the California Current System (above) and Principal Axes of Variance (below). Note that transport of drifters from the coastal region at 38°N to deep ocean at 30°N. (Courtesy of M. Swenson)
8.16 Canadian WOCE Surface Drifter Programme, Richard Thomson (Institute of Ocean Sciences, Canada)
Satellite-tracked drifters launched in the North Pacific Ocean since August 1990 are being used to investigate the scaling (fractal) properties, dispersion characteristics and eddy kinetic energy of currents in the upper Ocean. Flow separation of the Alaskan Stream, Kamchatka Current and cross-shelf exchange processes off the coast of British Columbia also are being investigated. A summary of these and other drifter-related research activities will be addressed.
Figure 13. Drifter Tracks in the Greenland Sea, June 1991 - February 1992. Stars mark deployment locations and circles mark last data point. Note the North Spitzberg Current is continuous across 36 degrees of longitude at 72°N latitude. (Courtesy of A. Warn-Varnas)
We have released surface drifters in the Iceland-Faeroe frontal region, in the vicinity of Greenland, in the Norwegian current, and off Iceland (Figure 13). The releases are part of our seasonal seeding programme in the GIN Sea. The tracks, temperatures, and instantaneous velocities have been plotted for some time periods. The tracks are related to the expected currents, frontal locations, eddies, and other features. Atmospheric data are being analyzed for storms in the drifter regions. Temperatures are interpreted with respect to satellite and our available field programme data, from summer 1991 and other cruises. An attempt to relate drifter deduced velocities with ADCP measurements, conducted by our ship Alliance, is in progress.
9.1 WOCE/SSG Requests for Information
WOCE/SSG has requested that:
(a) the analysis of the number of drifters in each ocean be indexed to how well these arrays resolve the WOCE 500 km scale circulation. GDC has agreed to provide this information as soon as possible.
(b) the Core Project 3 Working Group had decided that it was necessary during the next six months or so to re-examine the basis for the design of several elements of the experiment. The SVP PC agreed to evaluate, in the light of present knowledge, what would be obtained from the enhanced drifter coverage proposed for the North Atlantic. Professor Krauss agreed to chair a subgroup to carry out this task with the participation of Dr Richardson and other members from the oceanographic community.
9.2 Future WOCE Meeting Reports
SVP has been requested to make reports of their activities at:
(a) the April 1992 Core 2 meeting in La Jolla, California. Peter Niiler will attend that meeting and make the report.
(b) the May 1992 Core 1 workshop in the Pacific and Core 1 Planning Committee meeting. Dr Rick Thomson has agreed to make the appropriate reports.
SST measurements from SVP drifters are becoming more important in climate research (Section 6). It was deemed necessary to begin a long-term laboratory calibration test on the stability of the sensors and their electronic circuits as soon as possible. NDBC, under leadership of Ron Kozak, has agreed to work together with GDC to set up these tests and continue them for a several year period.
Prof Wolfgang Krauss' term of membership expires with this meeting. He will inform the Chairman shortly whether he will accept the request of SVP to continue his appointment for another three year period. Dr Haruo Ishii has left the Hydrographic Office of the Department of Transportation in Japan and he is replaced by Mr Y. Mishida as the director of the Japanese drifter programme. It was requested that Mr Y. Mishida replace Dr Ishii on the SVP Planning Committee. The term of R. Wilson from MEDS has expired. It is requested he continue on for another term or appoint an alternate from MEDS. Peter Niiler has accepted the WOCE/SSG request to continue on as Chairman of SVP Planning Committee for the next three years.
It was decided that SVP-6 be held in Honolulu in the Fall of 1993, i.e. less than a year before the WOCE/TOGA/JGOFS Scientific Meeting on the results of Pacific Ocean measurements and modelling.
9.6 Continuation of Drifter Measurements in GCOS and GOOS
A discussion was held in the executive session of SVP Planning Committee to consider the future of drifter observations into the onset of the GCOS and GOOS. TOGA observations will discontinue in 1995 and WOCE observations will discontinue in 1997. The SVP Planning Committee unanimously endorsed the concept of an expanded continuation of drifter measurements of surface velocity, SST and sea level atmospheric pressure within TOGA/WOCE/ACCP and an expansion of the measurements in the upcoming global ocean observing programmes. There are several planning committees for the design of the global ocean observing systems, and the members of SVP are encouraged to be aware of these committee deliberations and recommendations. From the earliest planning reports, it is quite apparent that both SST and surface velocity are important parameters to be measured for a long time in the global ocean. The numbers for required drifters, the levels of funding and the national commitments will be addressed at SVP-6, at which time also a more thorough discussion of the reports of these planning committees can be held.
WOCE/TOGA Surface Velocity Programme Planning Committee
SVP-5
6-8 April 1992
Hamilton, Bermuda
AGENDA
Monday, 6 April 1992
8:30 Welcome and Agenda Revisions
Reports of SVP Operational Centres
A 25 min update of the centre activities in the past year will be presented for each centre.
10:00 Break
10:30 Status and Survivability of Drifter Arrays
Presentations and discussions will be led by Niiler, Hansen and Sombardier.
12:00 Lunch
13:30 Update of National and Agency Programmes
Each country or agency will present a 10 min update of its drifter programme. Emphasis is on deployments in 1991/1992 and plans for 1993.
Tuesday 7, April 1992
8:30 Technical Developments
A detailed description of the latest developments will be given at this time.
09:30 Discussion of expanded temperature measurements from drifters (to be led by Reynolds)
10:00 Break
10:30 Discussion of expanded sea level pressure measurements from drifters (to be led by Niiler)
12:00 Lunch
13:30 New Science
Each scientific presentation is 30 min long. The presentations are listed here by authors: Allan, Beardsley, Du Penhoat, Figueroa, Otto, Krauss
Wednesday, 8 April 1992
8:30 New science
Maksimenko, Paduan, Poulain, Richardson, Swenson, Thomson, Warn-Varnas, Delecluse, Niiler (for McDowell), Niiler
10:00 Break
12:00 Lunch
15:30 WOCE/TOGA Planning Committee Executive Session
Bianchi, Bolduc, du Penhoat, Hansen, Kozak, Krauss, Maksimenko, Needler, Niiler, Painting, Partridge, Richardson, Sombardier, Szabados
PARTICIPANTS
|
Mr Art Allen |
Dr Robert Beardsley |
Dr Alejandro Bianchi |
|
Mr. André Bolduc |
Dr Etienne Charpentier |
Dr Pascale Delecluse |
|
Ms Valery Detemmerman |
Dr Yves du Penhoat |
Dr Horacio Figueroa |
|
Dr Donald Hansen |
Dr Jian-Hwa Hu |
Dr Ron Kozak |
|
Dr Wolfgang Krauss |
Dr Nikolay A. Maksimenko |
Dr Jack Murray |
|
Dr George Needler |
Dr Peter Niiler |
Dr Leo Otto |
|
Dr Jeffrey Paduan |
Dr Derek Painting |
Dr Ray Partridge |
|
Mayra Pazos |
Mr Berne Petrolas |
Dr Pierre-Marie Poulain |
|
Dr Richard Reynolds |
Dr Philip Richardson |
Dr Jean Rolland |
|
Mr Joseph Seiler |
Mrs Laurence Sombardier |
Dr Merritt Stevensen |
|
Dr Mark Swenson |
Mr Andrew Sybrandy |
Dr M. Szabados |
|
Dr Michel Taillade |
Dr Rick Thomson |
Dr Alex Warn-Varnas |
|
Mr Hank White |
Mr Gary Williams |
WOCE buoys have been tested and approved for rigging as a Container Delivery System (CDS) article for airdrop from C-130 aircraft. Following are rigging details required for airdrop preparation. Figure 14 shows a picture of the final rigging.
PRELIMINARY RIGGING
1. WOCE buoy will arrive shipped from factory wrapped in plastic. Remove buoy completely from plastic before continuing.
2. Place buoy hardware inside 48" square triwall military shipping box. Standard box is 48x48x36". Mark on top of the box the spot under which the transmitter magnet is located. This will permit later ease in turning transmitter on prior to airdrop.
3. Seal top of box with heavy adhesive tape around sides.
4. Take standard military wood shipping pallet (48" square) and secure 1/4" plywood sheet of plywood covering bottom of pallet. Secure plywood using wood screws.
*NOTE: In this configuration, buoy may be loaded aboard aircraft. Final airdrop rigging will be accomplished after aircraft is airborne on the airdrop mission. Plywood bottom to pallet permits ease of movement of buoy across C-130 floor roller system. Use of shipping pallet permits upload by forklift without direct contact of forklift to box.
**NOTE: The WOCE buoy will be air deployed from the C-130 ramp using the center set of floor rollers.
FINAL RIGGING
5. Using Type-8 nylon strapping, tie two parallel sections of Type-8 around the box following the route of the steel bands.
6. Attach one J-1 parachute to the sections of Type-8 nylon, connecting one attach point clip to each section. Center the parachute on the top of the box.
**NOTE: Parachute nomenclature:
J-1 Wind Drift Determination Parachute,7. At the attachment points of the parachute clip, tie one strand of 550 strength parachute riser cord, run beneath the box, and tie taught. This will maintain separation between the two sections of Type-8 during the air deployment.
**NOTE: Buoy in this configuration now weighs total 175 lbs.
8. Attach the static line to a floor ring on the ramp of the C-130 on the forward side of the box (direction nearest the nose of the aircraft).
9. Tie a strand of 500 strength parachute riser cord through the forward bottom openings in the shipping pallet and route through a floor ring on the forward (nearest the nose) side of the ramp. This will serve as a "gate" to be used to actually release the buoy.
10. Cut a hole in the top of the box at the place noted in Step 2 above. With your hand, remove the magnet on the flotation ball. This activates the buoy transmitter. Ensure the transmitter is functional by verification with appropriate test equipment.
11. Deploy the WOCE buoy from 500 feet absolute altitude while flying 130 knots indicated airspeed. Although deployment method is CDS, do not exceed 5 degrees deck angle. A flat deck angle (no inclination) is also effective. Excessive deck angle will cause the buoy to impact the tail of the aircraft after release.
12. The buoy will be physically deployed by cutting the gate (see Step 9) with a knife.
13. From the time the gate is cut to water impact from a drop altitude of 500 feet, expect the interval of 12 seconds. This includes approximately one second for the buoy to clear the ramp.
14. After water impact, the risers are sufficiently short on the J-1 parachute to be slammed into the water and sunk. The weight of the buoy cannot be completely floated by the amount of the wood in the rigging. Rather, the buoy will be buoyant, yet approximately one foot of the buoy's height will be submerged. This will allow saltwater degradation of the triwall box at a rapid rate. After approximately 30 minutes, the triwall box will be deformed sufficiently that the steel bands will fall away and there will be no visible sides to the box to hold the buoy inside. Within an hour after deployment, the buoy hardware once inside the box will slide off the wood pallet assisted by wave action. The drogue self deploys.
Figure 14. Schematic of the Air Deployment Configuration used by US Navy to deploy the WOCE/TOGA Lagrangian Drifters from cargo planes. (Courtesy of R. Partridge)