WMO IOC …

WMO                                                         IOC


                              JCOMM




  VOLUNTARY OBSERVING SHIPS (VOS) CLIMATE SUBSET PROJECT (VOSCLIM)




                         PROJECT DOCUMENT
                                    TABLE OF CONTENTS


                                                                                  page

1.     Objectives

2.     Strategy

3.     Required accuracy of data sets based on VOS observations

4.     Selection of ships

5.     Instrumentation and observations

6.     Data collection and verification

7.     Data management procedures

8.     Project management

9.     Information exchange


Attachment 1:         Scientific requirements and justification

Attachment 2:         Information required on first reconnaissance

Attachment 3:         Extra information with each observation

Attachment 4:         Terms of reference -- real time monitoring centre and DAC

Attachment 5:         List of focal points

Attachment 7:         Preliminary action plan
        PROJECT DOCUMENT FOR THE VOLUNTARY OBSERVING SHIPS (VOS)
                    CLIMATE SUBSET PROJECT (VOSCLIM)


1.     Objectives

         The primary objective of the project is to provide a high-quality subset of marine
meteorological data, with extensive associated metadata, to be available in both real time
and delayed mode. Eventually, it is expected that the project will transform into a long-term,
operational programme. Specifically, the project gives priority to the following parameters:
wind direction and speed, sea level pressure, sea surface temperature, air temperature and
humidity. Data from the project will be used: to input directly into air-sea flux computations, as
part of coupled atmosphere-ocean climate models; to provide ground truth for calibrating
satellite observations; and to provide a high-quality reference data set for possible re-
calibration of observations from the entire VOS fleet. Requirements, rationale and scientific
justification for the project are detailed in Attachment 1.

        The VSOP-NA demonstrated clearly that the quality of measurements depends
significantly on the types of instruments used, their exposures and the observing practices of
shipboard personnel. It made a number of substantive recommendations in these areas
which, if systematically implemented, would be expected to result in VOS observations of a
quality appropriate to global climate studies. For logistic reasons, it is not realistic to expect
full implementation to the entire global VOS. However, it is undoubtedly feasible for a limited
subset of the VOS, and the primary goal of this project is therefore to effect such a limited
implementation.

2.     Strategy

       VOSCLIM is intended to produce high-quality data and therefore the selection of
ships is a very important part of the project. This can best be done by the existing Port
Meteorological Officer (PMO) network applying the criteria detailed in section 4 below.

       Many modern ships will already be adequately equipped with acceptable instruments.
The quality of instruments, although relevant, has, however, less effect upon the quality of
data than use and exposure. It is vital that these aspects are catalogued and the catalogue
kept up to date.

        For data to be collected and processed in as timely a manner as possible, it is
necessary to implement new logsheets (or additions to existing logsheets), whether paper or
electronic, and additional codes. The new codes are to be used for parameters not currently
reported. Logsheets will be collected by PMOs of the recruiting country at the first
opportunity.

       Strenuous efforts must be made to ensure that the instruments, observing practice
and physical exposure conform to high standards as recommended in the VSOP-NA report,
and that details are accurately recorded. Intercomparison with reference instruments held by
PMOs could identify biases in specific instruments.

        The project necessarily involves the PMO network of participating countries. Close
liaison with the shipping industry, both companies and ships' masters, is essential for its
success. Expertise provided through the Joint WMO/IOC Technical Commission for
Oceanography and Marine Meteorology is also important and the project implementation
should therefore proceed as a component activity under, and as part of the integrated
observing network strategy of JCOMM. Overall scientific guidance for the project will be
provided by the Ocean Observations Panel for Climate (OOPC) of GOOS/GCOS/WCRP.
3.     Required accuracy of data sets based on VOS observations

         The most stringent accuracy requirement is that imposed by the need to calculate
heat, moisture and momentum flux values. These are required for validating the fluxes
calculated by atmospheric general circulation models and also by coupled ocean-atmosphere
models and for driving ocean circulation models. Flux values are also required for climate
research within WCRP. A similar level of accuracy is also imposed by the need to validate
satellite measurements.

        The precise accuracy depends on the spatial and temporal resolution for which data
are required. It also depends on the dominant physical processes, for example ocean mixed-
layer deepening over a month may be dominated by one short-term event. However, if one
considers for example the need to obtain a monthly mean value of the sensible and latent
heat fluxes over a 500 km square of ocean to about +/-10 W/m**2 (e.g. for climate research),
then the mean temperature measurements should be accurate to +/-0.2°C or better. This
applies to the dry and wet bulb air temperatures and to the sea surface temperature. An
equivalent requirement for the mean wind speed is about 10% or about 0.5 m/s (1 knot).

       It must be emphasized that individual ship observations are not expected to meet
such stringent accuracies. In any case there will always be random scatter from one
measurement to the next due to smaller-scale variability in the atmosphere and the ocean.
Thus is will be necessary to average sets of observations. However, if the random
observational errors are minimized, then the areas of ocean for which adequate densities of
observations are available will be increased. Systematic errors in the observations, on the
other hand, must be identified to the level of precision suggested above, so that any such
bias may be removed from the data sets.

4.     Selection of ships

        It is intended that only a relatively small number of ships (a target of 200) will be used
in the project. It is necessary to select those which make frequent and regular crossings of all
major ocean basins, in such a way as to provide a more or less global coverage in both
space and time. In addition, priority should be given to recruiting existing ASAP and /or
SOOP ships to the project, as well as ships sailing in the southern ocean, Antarctic supply
vessels and research ships, wherever feasible.

        The choice of individual ships is best left to national VOS personnel acting through
PMOs, but they must be provided with details of the ships' observing performance in the
past. The initial approach should be made to shore managers at the headquarters of likely
shipping lines. It will be necessary to prepare a short statement of the aims of the project and
long-term benefits to the shipping industry. Ship specifications (electronic drawings,
schematics, digital photographs, identification in catalogues) should be obtained by the
PMOs, together with the positions of instruments (see Attachment 2). PMOs will need to visit
individual ships to explain the project and obtain answers to the questionnaire in Attachment
3 and also to assess the likely commitment of observers to the task. Final selection of ships
will take note of existing instrumentation and exposure, past performance and the general
impression gained by the PMOs. Difficulties will arise where a ship is well-equipped for one
parameter but not another, e.g. no anemometer but mounting a hull sensor for sea surface
temperature. The value of the ship's contribution must be assessed in terms of the
importance of the parameters which are acceptable.

        PMOs will be required to explain the logbooks (electronic or paper), additional codes
and the method of completion. Later visits will also be necessary to check that instrument
exposure has not changed and discuss problems with observers. These observer "contact"
visits will be extremely important in maintaining the interest of the observers and the impetus
of the project. PMOs must be provided with "progress" material to leave with the observers
after visits, in addition to the briefing pack used on the first visit. At a later stage in project
development, consideration should be given to enhancing or upgrading instrumentation as
necessary, in line with VSOP-NA recommendations.

5.     Instrumentation and observations

          Ideally, ships taking part in the project should have the following instrumentation and
facilities:

(a)    Accurate and well-exposed thermometers with precision to 0.1C;
(b)    Sea surface temperature measuring instruments from hull contact sensors;
(c)    Permanently-mounted, well-exposed anemometers to 0.1 m/s precision;
(d)    Precision marine barometers to 0.1hPa precision, preferably connected to a static
       head;
(e)    Electronic logbook facility, to include true wind computation, QC checks and updated
       encoding in the revised code forms.

It is highly unlikely that common instrumentation will be achieved, but efforts should
nevertheless be made to implement those recommended in the VSOP-NA report. At the
same time, the most important characteristic of all instruments remains good exposure and
full documentation.

5.1    Instrument and data quality assessment

        The full intercomparison of instruments is a difficult and time-consuming task. It is
possible to compare typical samples of each type or source of manufacture, but often
variations between members of the same type are greater than between different
instruments. This problem will be addressed by intercomparison of the observations of the
VOS climate subset with large-scale fields or with neighbouring ships at sea. Comparisons
of real-time VOS weather observations with the predictions of Numerical Weather Prediction
(NWP) models are presently carried out routinely by several national weather centres around
the world. The monthly mean bias and scatter between individual ship observations and the
model fields are used to identify ships that are reporting poorly or have instruments that
require recalibration or replacement. This information is fed back to ships with significant
scatter or bias in their observations via the PMO network. If the co-located model values for
the basic meteorological variables were archived and passed to the data assembly centre
this would provide valuable validation information for the ship reports. Analysis of the VOS
climate subset could then be performed using a model as a comparison standard as in the
VSOP-NA project. Whilst it is envisaged that NWP model data merged with individual ship
reports will form the main source of information on bias and scatter for a particular ship or
measurement technique, satellite data may now also form a valuable potential source of
background field for certain variables. Comparison of ship observations with nearby reports
from other ships is another source of validation information.

        Regular checks upon the serviceability of instruments by the PMOs are essential.
Calibration of temperature sensors is best performed using a water bath and it is possible to
calibrate some types of wind speed sensor by mechanically rotating the propeller.
Documentation of calibrations and comparisons of instruments made on visits to the ships is
necessary.

5.2    Observations

        The normal SHIP observation will be retained, augmented by extra code groups for
additional parameters. This extra information is essential to the success of the project.
       The following additional information will be needed (codes quoted in Attachment 3).

Ship's parameters

-Ship's speed over ground at time of observation                                     Code 1
-Ship's heading at time of observation                                               Code 2
-Height of deck cargo above summer maximum load line                                 Code 3
-Departure of summer maximum load line from actual sea level                         Code 4

Wind (for ships with anemometers)

-Uncorrected (i.e. relative) speed and direction                                     Code 5
                                                                                     Code 6


6.     Data collection and verification

        It is intended that the project should provide timely and complete information. To this
end, and to ensure that no reports (or information contained in these reports) from
participating ships are lost, data will be submitted in both real time and delayed mode. Real
time data submission will be in the form of reports on the GTS in the SHIP code. In support of
the project, the standard SHIP code will be augmented with additional optional groups in
Section 2, as detailed in Attachment 3. Reports from participating ships will be inserted onto
the GTS in the normal manner. It is the responsibility of the Data Assembly Centre (see
Section 7.2 below) to identify and extract these reports on the basis of a published and
continuously updated list of call signs of participating ships.

         In addition, all observations will be recorded for delayed mode submission in either
paper or electronic logbooks. Paper logbooks will be specially designed for the project, and
will include all the additional information transmitted with the real time reports. The recruiting
country will digitise these observations in revised IMMT format, apply the agreed minimum
QC procedures, and forward the digital data sets to the Data Assembly Centre (on diskette or
via internet) with a minimum delay. Electronic logbooks will necessitate some extension to
existing procedures (e.g. TurboWIN). When these logbooks are collected by recruiting
countries, they should be carefully screened for duplicates before full data sets are compiled
and forwarded as detailed above.

        The project will also require real time observational data monitoring, and comparison
with model fields. To this end, a Real Time Monitoring Centre has been established by the
U.K. Meteorological Office (which already undertakes such monitoring of ship observations
on a routine basis). Observations from participating ships will be identified by this centre, and
associated with co-located model field values. These data sets will also be transferred on a
regular basis to the Data Assembly Centre. Terms of reference for this monitoring centre are
given in Attachment 4.

      Note that all observations under the project should also continue to be sent to the
GCCs in the normal manner, as part of the MCSS.


7.     Data Management procedures

7.1    Instrument exposure and other metadata

       To achieve the accuracies described in section 3, it is vital to have information about
the exposure of instruments and the actions necessary to allow for that exposure, as well as
other metadata as detailed in the VSOP-NA recommendations. Details of metadata, including
digital imagery, instrument exposure and date of any changes, and ratings of the quality of
instrument exposures, will be stored in a master index of ships. This will be developed as a
supplement to, but separate from the main ship catalogue (WMO-No. 47). The catalogue will
be continuously updated and made available through the Data Assembly Centre, as well as
distributed in hard copy form (loose-leaf) to all PMOs through the national contact points.
This catalogue will contain details of the instrument locations for each ship in an agreed
format. The catalogue will also contain details of the results of regular ship inspections by
PMOs.

7.2    Data assembly

        The data collected during the project will be collected, quality controlled and archived
by the project Data Assembly Centre. The data assembly centre will create and maintain a
relational database so that the information on instrument types, exposure and observing
practice can be automatically associated with each observation. The database should be
freely accessible to registered users. Terms of reference for the data assembly centre are
given in Attachment 4. The National Climatic Data Center, NOAA, USA, will undertake this
task.

8.     Project management

        The nature and scope of the project necessitate the involvement of a number of
national Meteorological Services which have recruited substantial numbers of VOS operating
worldwide. In particular the active participation at least of Argentina, Australia, Canada,
China, France, Germany, Japan, Netherlands, New Zealand, United Kingdom and the USA is
essential to the success of the project. Overall management of the project will be undertaken
by a Management Group, comprising representatives of all participating countries and the
OOPC. The Management Group will also require the involvement/support of the WMO
Secretariat in discharging its functions. This group will have responsibility for project
supervision, development of specific activities and sub-projects, allocation of responsibilities
for action, review of progress, preparation of periodic reports including recommendations,
and deciding on any other actions as may be necessary for the successful continuation of the
project. The chairman of the Management Group will also be Project Leader, with
responsibilities to act as a focal point for the project and on behalf of the Management Group,
as necessary, as well as to undertake other tasks as indicated in this Project Document.

        Members of the Management Group will also act as national focal points for the
project, with responsibility for implementation of the project at the national level, for reporting
to the Project Leader as appropriate, for agreeing to project modifications, as necessary, and
for undertaking other tasks as given in this Project Document, in particular the sub-projects
allocated to specific national Meteorological Services. From the national focal points,
responsibilities for action will devolve to PMOs and others within national Meteorological
Services and ultimately to the ships' masters. The list of project focal points is given in
Attachment 5.

        Following the initial implementation co-ordination meeting, most of the project
organization and implementation may be achieved by correspondence. However, some
further meetings will be necessary, to assess progress at appropriate intermediate stages
and, if necessary, to refine various aspects of the project. Decisions will also eventually be
required on transforming the project into an ongoing programme. An initial action plan for the
project is given in Attachment 6.

9.     Information exchange

       Extensive and frequent information exchange is an essential aspect of the project.
This exchange must take place among participants (including the focal points, PMOs and
ships crews) and with users. A primary means for such exchange will be a special project
web site, implemented and maintained by the DAC, with contributions to be made by both
participants and users (see terms of reference for the DAC in Attachment 4). Information to
be made available through the web site will include:

·      metadata catalogue of participating ships
·      regular project update reports
·      monitoring and data application results
·      data catalogues
·      a project newsletter for participating ships
·      links to other relevant web sites
·      project focal points and other relevant contact details
·      the project document and other publications
·      any other information relevant to the project

        The project newsletter for participating ships is an important means of regularly
informing ships crews of the status of the project in general and of their own specific
contributions, as well as of the applications of project data. It should thereby help to maintain
interest and enthusiasm among these crews. Contributions to the newsletter should come
from all participating ship operators, the DAC and monitoring centre, users and ships crews
themselves whenever possible. These contributions should be sent to the project leader and
Secretariat so as to allow preparation of a newsletter at least every six months (January and
July). The newsletter will be transmitted by the Secretariat in electronic form to the DAC,
where it will be made available on the web site in a suitable format to allow downloading by
participating operators for printing and distribution to ships.

                                              oOo
                                                                                Attachment 1

                         Scientific requirements and justification

1. The evolving requirements for Voluntary Observing Ship data

1.1 Introduction


        For well over 100 years, the weather observations from merchant ships have been
used to define our knowledge of the marine climate. This function continues within the
Voluntary Observing Ships (VOS) programme as the Marine Climatological Summaries
Scheme. However the main emphasis of the VOS programme has traditionally been the
provision of data required for atmospheric weather forecasting. Today, the initialisation of
numerical weather prediction models remains an important use of weather reports from the
VOS. However recent trends, such as the increasing availability of data from satellite
sensors, and the increased concern with regard to climate analysis and prediction, are
making further requirements for data from the VOS.


       That there is a growing need for higher quality data from a sub-set of the VOS has
been identified by, inter alia, the Ocean Observing System Development Panel (OOSDP,
1995), the Ocean Observations Panel for Climate (OOPC, 1998), and the JSC/SCOR
Working Group on Air Sea Fluxes (WGASF, 2000). The justification for improved surface
meteorological data was also discussed in detail at the recent Conference on the Ocean
Observing System for Climate (see paper by Taylor et al. 1999). Here we shall give
examples of the requirements, the present state of the art and the potential improvements.


1.2 Examples of evolving requirements for VOS data


A. SATELLITE DATA VERIFICATION
         Satellite borne sensors are now used routinely for, for example, determining sea
surface temperature (SST), sea waves, and surface wind velocity. Compared to in situ
measurements, these remotely sensed data provide better spatial coverage of the global
oceans. However the data are derived from empirical algorithms and a very limited number
of individual sensors. In this respect, an important role for VOS data is the detection of
biases in the remote sensed data due to instrument calibration changes or changing
atmospheric transmission conditions. For example, the SST analyses produced by the US
National Centers for Environmental Prediction (NCEP) are used at a number of operational
weather forecasting centres including the ECMWF. The NCEP analyses (Reynolds and
Smith, 1994) use SST data from satellite sensors that have been initially calibrated against
drifting buoy data. VOS and buoy data are used to detect and correct biases in the satellite
data caused, for example, by varying atmospheric aerosol loading due to volcanic eruptions.
Without these real time bias corrections, errors of several tenths K or more can occur in
satellite derived SST values (Reynolds, 1999). For satellite verification purposes the need is
for a dataset of accurate data with known error characteristics.


B. CLIMATE CHANGE STUDIES
        The VOS data are being increasingly used for climate change studies. Assembled
into large data bases (such as the Comprehensive Ocean Atmosphere Data Set, COADS,
Woodruff et al., 1993) the observations have been used, for example, to quantify global
changes of sea and marine air temperature (Folland and Parker, 1995). Based on such
studies, the recommendations of the Intergovernmental Panel on Climate Change
(Houghton et al., 1990) have led to politically important international agreements such as the
UN Framework Convention on Climate Change. However the detection of climate trends in
the VOS data has only been possible following the careful correction, as far as is possible, of
varying observational bias due to the changing methods of observation. For example sea
temperature data have different bias errors depending on whether they were obtained using
wooden buckets from sailing ships, canvas buckets from small steam ships, or engine room
intake thermometers on large container ships. For the present, and for the future, it is
important that we better document the observing practices that are used.


C. CLIMATE RESEARCH AND CLIMATE PREDICTION
         Coupled numerical models of the atmosphere and ocean are increasingly being used
for climate research and climate change prediction. Because the time and space scales for
circulation features in the atmosphere and the ocean are very different, the ocean surface is
an important interface for model verification. The simulated air-sea fluxes of heat, water
and momentum must be shown to be realistic if there is to be confidence in the model
predictions. At present the uncertainty in our knowledge of these surface fluxes is of a
similar order to the spread in the model predictions (WGASF, 2000). Partly this is due to the
limitations of the parameterisation formulae used to calculate the fluxes. Verification of the
model predictions of near surface meteorological variables (air temperature, humidity, SST
etc.) against high quality in situ observations from moored "flux" buoys and specially
selected VOS is required (e.g. Send et al. 1999, Taylor et al. 1999a).

2. The State-of-the-Art for VOS observations

2.1 What is needed?


        These relatively new applications for VOS data imply a need to minimise the errors
present in the observations. For example, 10 Wm-2 is often quoted as a target accuracy for
determining the heat fluxes; it is about 10% of the typical interannual variability of the
wintertime turbulent heat fluxes in mid to high latitudes. To achieve such accuracy implies
that the basic meteorological fields are known to about ­0.2°C for the SST, dry and wet bulb
temperatures (or about 0.3 g/kg for specific humidity) and that the winds be estimated to
­10% or better, say about 0.5 m/s. These are stringent requirements which we do not
expect to be met by an individual VOS observation.              Enough observations must be
averaged to reduce the errors to the required level. The more accurate the individual VOS
observations, the less averaging will be needed.            Nor is averaging alone enough;
corrections must also be applied for the systematic errors in the data set.


        In terms of the longer term ocean heat balance even an accuracy of 10 Wm-2 is not
adequate. A flux of 10 Wm-2 over one year would, if stored in the top 500m of the ocean,
heat that entire layer by about 0.15°C. Temperature changes on a decadal timescale are at
most a few tenths of a degree (e.g. Parilla et al., 1994) so the global mean heat budget must
balance to better than a few Wm-2. It is unlikely that such accuracy will ever be achieved
using VOS data either alone, or combined with other data sources. Thus the calculated flux
fields must be adjusted, using "inverse analysis", to satisfy various integral constraints.
Inverse analysis techniques rely on detailed knowledge of the error characteristics of the
data; information which is poorly known at present for the VOS data set. Thus there is an
urgent need to better define the accuracy of VOS data.
2.2 What is presently achieved?


        To attempt to quantify the random error in VOS observations, Kent et al. (1999)
determined the root-mean-square (rms) error for VOS reports of the basic meteorological
variables. Table 1 shows the minimum, maximum and mean error values for individual ship
observations calculated for 30° x 30° areas of the global ocean. It is obvious that individual
ship observations can not achieve the desired accuracy and that the average of many
observations is needed. For example, to reduce a typical temperature error of 1.4C to the
desired 0.2C requires some 50 independent observations; more when natural variability is
taken into account. Sufficient observations are obtained for adequate monthly mean values
in well-sampled regions like the North Atlantic but in data sparse regions acceptable
accuracy cannot be achieved.


        The Voluntary Observing Ship Special Observing Programme - North Atlantic project,
VSOP-NA (Kent et al., 1993a), was an attempt to determine the systematic errors in VOS
data. For a subset of 46 VOS, the instrumentation used was documented (Kent & Taylor,
1991), and extra information included with each report. The output from an atmospheric
forecast model was used as a common standard for comparison.           The results were
analysed according to instrument type and exposure, ship size and nationality and other
factors, and relative biases were determined. For example it was found that SST values
from engine intake thermometers were biased warm compared to other methods (Kent et al.
1993a), and that daytime air temperatures were too warm due to solar heating (Kent et al.
1993b). It could be shown that the dew point temperature was not biased by the latter error
(Kent and Taylor, 1996) but, compared to aspirated psychrometer readings, the dew point
was biased high when obtained from fixed thermometer screens.


     Table 1 - RMS Error Estimates: The uncertainty quoted in the mean error is
     derived from the weighted sum of the error variances (from Kent et al. 1999)

               Observed Field                            RMS Error:
                                                  Min.      Max.       Mean
               Surface Wind Speed (m/s)            1.3       2.8      2.1 ± 0.2
               Pressure (mb)                       1.2       7.1      2.3 ± 0.2
               Air Temperature (°C)                0.8       3.3      1.4 ± 0.1
               Sea Surface Temperature             0.4       2.8      1.5 ± 0.1
               (°C)
               Specific Humidity (g/kg)            0.6       1.8      1.1 ± 0.2




        Some of the VOS in the VSOP-NA project reported anemometer estimated, relative
wind speed in addition to the calculated true wind speed. Kent et al. (1991) showed that a
major cause of error was the calculation of the true wind speed. Only 50% of the reported
winds were within 1 m/s of the correct value, 30% of the reports were more than 2.5 m/s
incorrect. For wind direction, only 70% were within ±10° of the correct direction and 13%
were outside ±50°. These were large, needless errors which significantly degraded the
quality of the anemometer winds. A similar conclusion was reached by Gulev (1999).
Preliminary results from a questionnaire distributed to 300 ships' officers showed that only
27% of them used the correct method to compute true wind. The problem is not confined to
VOS observations. A majority of the wind data sets obtained from research ships during the
World Ocean Circulation Experiment showed errors in obtaining true wind values (Smith et
al., 1999).


2.3 How can the situation be improved?


         Consider as an example, wind velocity. The typical rms error for a wind speed
observation shown in Table 1 (about 2.1 m/s) was achieved after instrumental observations
had been corrected for the height of the anemometer above the sea surface (using data from
the List of Selected Ships, "WMO47") and the visual observations corrected using the Lindau
(1995) version of the Beaufort scale. For the observations as reported, the errors were
nearly 20% greater - about 2.5 m/s. Alone, this change in mean accuracy decreases the
number of observations required to obtain a reliable mean by a factor of 2/3rds. The quality
of the anemometer winds can be further improved by using an automated method of true
wind calculation such as the TurboWin system developed at KNMI. The effect on the
anemometer measurements of the air-flow disturbance around the ships' hull and
superstructure can be investigated using computational fluid dynamics (CFD) modelling of
the airflow (Yelland et al., 1998). While it would be impracticable to model all the VOS, it is
believed that typical values for the resulting error can be estimated given knowledge of the
anemometer position and the overall geometry of the ship (Taylor et al. 1999b).


        Similarly for the other observed variables correction schemes can be devised. For
example, air temperature errors due to daytime heating of the ship depend on the solar
radiation and the relative wind speed (Kent et al. 1993b). Josey et al., (1999) found that
correcting the various known biases changed the climatological monthly mean heat flux by
around ­15 Wm-2 varying with area and season. For climate studies these represent
significant changes.

3. Conclusions

         Most of the potential improvements discussed above require detailed, accurate
documentation on the methods of observation. Some of this information is available in the
List of Selected Ships (WMO47) which should be augmented with information similar to that
collected for the ships which participated in the VSOP-NA. Improved meta-data with regard
to the ship and observing practices, and improved quality control of the observations, are the
initial priorities for the VOS Climate project. Other desirable enhancements to the VOS
system include increased use of automatic coding, and improved instrumentation. These
are being introduced on an increasing number of VOS, and future implementation on the
ships participating in the VOS climate subset should be anticipated.


       The successful implementation of the VOS Climate project will represent an
important contribution to the Ocean Observing System for Climate as defined by the OOSDP
(1995) and the OOPC (1998).

4. References

Folland, C. K. and D. E. Parker, 1995: Correction of Instrumental Biases In Historical Sea-
       Surface Temperature Data. Q. J. Roy. Met. Soc., 121(522), 319 - 367.
Gulev, S. K., 1999: Comparison of COADS Release 1a winds with instrumental
      measurements in the Northwest Atlantic. J. Atmos. & Oceanic Tech., 16(1), 133 -
      145.
Houghton, J. T., G. J. Jenkins and J. J. Ephraums, 1990: Climate Change: The IPCC
      Scientific Assessment. , Cambridge University Press, 365 pp.
Josey, S. A., E. C. Kent and P. K. Taylor, 1999: New insights into the Ocean Heat Budget
       Closure Problem from analysis of the SOC Air-Sea Flux Climatology. Journal of
       Climate, 12(9), 2856 - 2880.
Kent, E. C. and P. K. Taylor, 1991: Ships observing marine climate: a catalogue of the
       Voluntary Observing Ships Participating in the VSOP-NA. Marine Meteorology and
       Related Oceanographic Activities 25, World Meteorological Organisation, Geneva,
       123 pp.
Kent, E. C. and P. K. Taylor, 1996: Accuracy of humidity measurements on ships:
      consideration of solar radiation effects. J. Atmos. & Oceanic Tech., 13(6), 1317 -
      1321.
Kent, E. C., B. S. Truscott, J. S. Hopkins and P. K. Taylor, 1991: The accuracy of ship's
       meteorological observation - results of the VSOP-NA. Marine Meteorology and
       Related Oceanographic Activities 26, World Meteorological Organisation, Geneva, 86
       pp.
Kent, E. C., P. Challenor and P. Taylor, 1999: A Statistical Determination of the random
       errors present in VOS Meteorological reports. J. Atmos. & Oceanic Tech., 16(7), 905
       - 914.
Kent, E. C., P. K. Taylor, B. S. Truscott and J. A. Hopkins, 1993a: The accuracy of Voluntary
       Observing Ship s Meteorological Observations. J. Atmos. & Oceanic Tech., 10(4),
       591 - 608.
Kent, E. C., R. J. Tiddy and P. K. Taylor, 1993b: Correction of marine daytime air
      temperature observations for radiation effects. J. Atmos. & Oceanic Tech., 10(6), 900
      - 906.
Lindau, R., 1995: A new Beaufort equivalent scale. Internat. COADS Winds Workshop, Kiel,
       Germany, 31 May - 2 June 1994, Environmental Research Labs., National Oceanic
       and Atmospheric Administration, Boulder, Colorado, 232 - 252.
OOPC (1998) Report of the Third Session of the Joint GCOS-GOOS-WCRP Ocean
     Observations Panel for Climate (OOPC), Grasse, France, 6 - 8 April 1998, GOOS
     Report No. 61, GCOS Report No. 44, IOC/UNESCO Paris, 37pp. + Annexes.
OOSDP (1995) Scientific Design for the Common Module of the Global Ocean Observing
     System and the Global Climate Observing System: An Ocean Observing System for
     Climate, Dept. of Oceanography, Texas A&M University, College Station, Texas,
     265 pp.
Parilla, G., Lavin, A., Bryden, H., Garcia, M. and Millard, R. (1984) Rising temperature in the
         subtropical North Atlantic Ocean over the past 35 years, Nature, 369, 48 - 51.
Reynolds, R. W. (1999) Specific contributions to the observing system: Sea Surface
      Temperatures. Proc Conf. The Ocean Observing System for Climate - Oceanobs 99,
      St Raphael, France, 25 - 27 October, 1999.
Reynolds, R. W. and T. M. Smith, 1994: Improved global sea surface temperature analyses
      using optimum interpolation. J. Climatol., 7, 929 - 948.
Send, U., Weller, R. A., Cunningham, S., Eriksen, C., Dickey, T., Kawabe, M., Lukas, R.,
      McCartney, M. and Osterhus, S. (1999) Oceanographic time series observatories.
      Proc Conf. The Ocean Observing System for Climate - Oceanobs 99, St Raphael,
      France 25 - 27 October, 1999.
Smith, S. R., M. A. Bourassa and R. J. Sharp, 1999: Establishing more truth in true winds. J.
       Atmos. & Oceanic Tech., 16(7), 932 - 952.
Taylor, P. K., Bradley, E. F., Fairall, C. W., Legler, L., Schulz, J., Weller, R. A. and White, G.
        H. (1999) Surface Fluxes and Surface Reference Sites. Proc Conf. The Ocean
        Observing System for Climate - Oceanobs 99, St Raphael, France, 25 - 27 October,
        1999.
Taylor, P. K., Kent, E. C., Yelland, M. J. and Moat, B. I. (1999) The Accuracy Of Marine
       Surface Winds From Ships And Buoys. CLIMAR 99, WMO Workshop on Advances in
       Marine Climatology, Vancouver, Canada, 8 - 15 Sept. 1999, 59 - 68.
WGASF, 2000: Intercomparison and validation of Ocean-Atmosphere energy flux fields,
    Final Report of the WDRP/SCOR Working Group on Air-Sea Fluxes (SCOR Working
    Group 110) various pagination.
Woodruff, S. D., S. J. Lubker, K. Wolter, S. J. Worley and J. D. Elms, 1993: Comprehensive
      Ocean-Atmosphere Data Set (COADS) release 1a: 1980-92. Earth System Monitor,
      4(1), 4 - 8.
Yelland, M. J., B. I. Moat, P. K. Taylor, R. W. Pascal, J. Hutchings and V. C. Cornell, 1998:
       Measurements of the open ocean drag coefficient corrected for air flow disturbance
       by the ship. J. Phys. Oceanogr., 28(7), 1511 - 1526.
                                                                                    Attachment 2

                  INFORMATION REQUIRED ON FIRST RECONNAISSANCE

This information will be obtained by PMOs and, for those ships selected, be forwarded to the
Data Assembly Centre for compilation of the metadata catalogue, to be made available to all
participating services.

Air temperature

(a)    Type of screen or housing;
(b)    Type of thermometer and make;
(c)    Method of estimating humidity, giving type and make of instrument used (could
       include sling psychrometer);
(d)    Measurement location of thermometers onboard ship.

Sea-surface temperature

(a)    Method of measurement, e.g. bucket, engine room intake, hull sensor, trailing
       thermistor or any other;
(b)    Type of thermometer used;
(c)    For non-bucket, location of instrument, e.g. distance inboard or depth below reference
       level;
(d)    If a bucket is used, details of type and insulation;
(e)    For bucket, agreed location of measurement. Also need location of nearby engine
       room and other discharges.

Wind

(a)    Type of anemometer used, including whether fixed or portable;
(b)    Location of anemometer and height above reference level;
(c)    Averaging time used for wind speed;
(d)    Units used in measuring wind speed;
(e)    Is a correction applied for ship's speed and direction: if yes give details of method;
(f)    How is the ship's instantaneous speed measured.

Pressure

(a)    Type of barometer/barograph;
(b)    Location and height above reference level;
(c)    Height of pressure head above reference level (if P.H. fitted).

NOTES: (1)        The location of instruments should be marked on ship's drawings, in
                  addition to detailed written descriptions and photographs (JPEG) of such
                  locations by the PMOs. The descriptions should also give details of any
                  permanent structural features which will affect the observation, e.g. water
                  outfalls, airflow obstructions, air conditioning vents, etc.
           (2)    The reference level to be used will be the summer maximum load line.
                                                                            Attachment 3

                EXTRA INFORMATION WITH EACH OBSERVATION

Ship parameters

Code 1            ss          Instantaneous ship's speed in knots at time of
                              observation
Code 2            DD          Ship's heading in tens of degrees true
Code 3            LL          Maximum height in metres of deck cargo above summer
                              maximum load line
Code 4            hh          Departure of summer maximum load line from actual
                              sea level (m)

Wind

Code 5            ff          Relative speed in knots or m/s (in conformity with wind
                              code indicator)
Code 6            DD          Relative wind direction in tens of degrees (00 to 36) off
                              the bow.

This information will be included in Section 2 of the SHIP code, in optional groups to
be introduced after the ICE groups. The groups will be prefixed by CLIM, and will be
of the form 1ssDD 2LLhh 3ffDD, to be extended as required.
                                                                      Attachment 4

                      REAL TIME MONITORING CENTRE

                              Terms of Reference

1.   Extract GTS reports of project ships (by call sign) and decode (including
     additional project data).
2.   Associate project observed variables (pressure, air temperature, humidity,
     SST, wind speed and direction) for each project ship with co-located model
     field values (4 times daily).
3.   Compile data sets of observations and associated model field values and
     transfer to the Data Assembly Centre (daily/monthly).
4.   Provide existing ship monitoring statistics for all VOS to Data Assembly
     Centre.

                                      oOo

                         DATA ASSEMBLY CENTRE

                              Terms of Reference

1.   Extract GTS reports of project ships (by call sign) and decode (including
     additional project data).
2.   Receive the real time reports from the Real Time Monitoring Centre
3.   Collect delayed mode reports of project ships from participants.
4.   Merge real time and delayed mode reports, eliminate duplicates and compile
     a complete project data set.
5.   Collect metadata and survey reports for project ships and compile a complete
     data set.
6.   Make project data sets available to users on request.
7.   Maintain a project web site, information exchange mechanism and electronic
     newsletter.


                                      oOo
                                                                         Attachment 5
                                  LIST OF FOCAL POINTS


PROJECT LEADER                                E-mail: songlc@sky.nmc.gov.cn

Capt. G. Mackie                               FRANCE
WMO marine consultant
30 Keephatch Road                             Mr Michel Hontarrède
WOKINGHAM RG40 1QJ                            JCOMM Rapporteur on Graphics
United Kingdom                                Transmission
Tel: + 44 1189 783 687                        Météo-France
Fax: + 44 1189 890 379                        D2C/COM/PUB
Email: gvmackie@cs.com                        1 quai Branly
                                              75340 PARIS
ARGENTINA                                     France
                                              Telephone: +33-1 45 56 74 60
Licenciada Ana Teresa Gomez                   Telefax: +33-1 45 56 74 59
Servicio Meteorologico Nacional               E-mail: michel.hontarrede@meteo.fr
25 de Mayo 658
1002 BUENOS AIRES                             GERMANY
Argentina
Telephone: +54-11 4514 4221                   Dr Volker Wagner
Telefax: +54-11 4514 4225                     Deutscher Wetterdienst
E-mail: cim@meteofa.mil.ar                    Geschäftsfeld Seeschiffahrt
                                              Postfach 30 11 90
AUSTRALIA                                     D-20304 HAMBURG
                                              Germany
Mr David K. Evans                             Telephone: +49-40 3190 8821
Bureau of Meteorology                         Telefax: +49-40 3190 8941
150 Lonsdale Street                           E-mail: volker.wagner@dwd.de
MELBOURNE, Vic. 3000
Australia                                     JAPAN
Telephone: +61-3 9669 4205
Telefax: +61-3 9669 4168                      Ms Teruko Manabe
E-mail: d.evans@bom.gov.au                    Maritime Meteorological Division
                                              Climate and Marine Department
CANADA                                        Japan Meteorological Agency
                                              1-3-4 Otemachi, Chiyoda-ku
Mr Ron Fordyce                                TOKYO 100-8122
Supt. Marine Data Unit                        Japan
Environment Canada                            Telephone: +81-3 3212 8341 ext. 5163
Port Meteorological Office                    Telefax: +81-3 3211 6908
100 East Port Blvd                            E-mail: teruko.manabe@met.kishou.go.jp
HAMILTON, Ontario L8H 7S4
Canada                                        NETHERLANDS
Telephone: +1-905 312 0900
Telefax: +1-905 312 0730                      Mr F.B. Koek
E-mail: Ron.Fordyce@ec.gc.ca                  Royal Netherlands Meteorological
                                               Institute (KNMI)
CHINA                                         P.O. Box 201
                                              3730 AE DE BILT
Mr Song Lianchun                              Netherlands
China Meteorological Administration           Telephone: +31-30 2206860
46 Baishiqiao Road                            Telefax: +31-30 2204614
BEIJING 100081                                E-mail: koek@knmi.nl
China
Telephone: +86-10 6840 6325                   NEW ZEALAND
Telefax: +86-10 6217 4797
Telex: 22094 FDSMA CN                         Ms Julie Fletcher
Marine Meteorological Officer         United Kingdom
Meteorological Service of NZ Ltd.     Tel: + 44 2380 596 408
PO Box 722                            Fax: + 44 2380 596 400
WELLINGTON                            Email: peter.k.taylor@soc.soton.ac.uk
New Zealand
Telephone: + 644 4700 789             Ms Elizabeth C. Kent
Telefax: + 644 4700 772               James Rennell Division (254/31)
Email: fletcher@met.co.nz             Southampton Oceanography Centre
                                      SOUTHAMPTON SO14 3ZH
UNITED KINGDOM                        United Kingdom
                                      Tel: + 44 2380 596 409
Ms Sarah North                        Fax: + 44 2380 596 400
Meteorological Office                 E-mail: elizabeth.c.kent@soc.soton.ac.uk
Observations Voluntary Branch
Scott Building, Eastern Road          WMO SECRETARIAT
BRACKNELL, Berkshire RG12 2PW
United Kingdom                        Dr P.E. Dexter
Tel: + 44 1344 855 652                World Weather Watch Department
Telefax: +44-1344 855 921             World Meteorological Organization
Email: snorth@meto.gov.uk             7 bis, Avenue de la Paix
Telex: 849801 WEAKBA G                Case postale No 2300
                                      CH-1211 GENEVE 2
Captain Edward O'Sullivan             Switzerland
Meteorological Office                 Telephone: +41-22 730 82 37
Observations Voluntary Branch         Telefax: +41-22 730 80 21
London Road                           E-mail: dexter@www.wmo.ch
BRACKNELL, Berkshire RG12 2SZ
United Kingdom
Telephone: +44-1344 855 913
Telefax: +44-1344 855 921
E-mail: ejosullivan@meto.gov.uk
Telex: 849801 WEAKBA G

USA

Mr Joe D. Elms
National Climatic Data Center
151 Patton Avenue
ASHEVILLE, NC 28801-5001
USA
Telephone: +1-828 271 4436
Telefax: +1-828 271 4328
E-mail: jelms@ncdc.noaa.gov

Mr Vincent Zegowitz
NOAA/NWS
1325 East-West Highway, Room 14112
SILVER SPRING, Maryland 20910
USA
Telephone: +1-301 713 1677 ext. 129
Telefax: +1-301 713 1598
E-mail: Vincent.Zegowitz@noaa.gov

OCEAN OBSERVATIONS PANEL FOR
CLIMATE

Dr Peter K. Taylor
James Rennell Division (254/27)
Southampton Oceanography Centre
SOUTHAMPTON SO14 3ZH
                                                                             Attachment 6

                                 PRELIMINARY ACTION PLAN

                  ACTION                                     WHOM               WHEN

1. Agree project leader and national focal    Meeting                        Done
   points


2. Identify and agree Real Time               UKMO/NCEP                      February
   Monitoring Centre/terms of reference                                      2000


3. Identify and agree Data Assembly           NCDC/DWD/UKMO                  February
   Centre/terms of reference                                                 2000


4. Implement agreed SHIP code change          WMO/CBS                        6 months


5. Implement agreed IMMT format               WMO/JCOMM                      April 2000
   change


6. Develop and implement required             KNMI/USA, participants, WMO    6 months
   electronic and paper log changes                                          (dependant
                                                                             on 4 and 5)

7. Develop, agree, distribute draft project   P. Taylor, V. Zegowitz, WMO    March 2000
   promotional material                       with participants


8. Finalize project plan                      WMO, Project Leader            February
                                                                             2000

9. Design/agree ship survey/ship              SOC, BOM, NWS, Project         March 2000
   inspection report form                     Leader, with participants


10. Design Metadata catalogue (No. 47)        WMO, DAC, JCOMM                August 2000
   supplement


11. Design ship award, project name and       V. Zegowitz, Project Leader,   July 2000
   project logo                               WMO, participants


12. Undertake preliminary ship                Participants                   Before next
   identification                                                            meeting


13. Convene second project meeting            WMO, Project Leader, DAC to    September/
                                              host                           October 2000