The development of Turbo (later renamed TurboWin) has always been a task that was taken by KNMI without obligations. In the first place it was developed for our own national use, but since other countries were interested in working this same way, a broader approach was set in. The advantage of this in-house development is the flexibility and this has proven worthwhile already several times. Working together with and for other WMO-members demanded a more indirect approach. We receive many remarks, comments, suggestions and etc. from our colleagues and from the observers, which is most valuable. Nevertheless, given all the restrictions and boundaries that WMO regulations imply on us, we had to make decisions in some matters. One of those items I refer to is the utilization of a height reduction factor to the measured wind speed on board the VOS.

WMO regulations (e.g. WMO-Guide to meteorological instruments and methods of observation, WMO No. 8, 6^th ed., WMO, Geneva, 1996) advise that wind should be reported at a height of 10m. Before the arrival of anemometers on board of ships the wind speed was normally estimated visually, after the appearance of the waves. Present-day observation practice on board still uses this method in many cases. The use of anemometers was initially discouraged because of the poor exposure. The use of anemometers on offshore rigs and platforms marked the beginning of the real discussion on this topic. Observers on these platforms were not able to make a good visual estimation of the waves (an consequently gave a wrong wind speed) so they were “allowed” to use anemometers. Since these instruments were frequently placed in derricks and masts, easily higher than 50m above sea level, there was a need to reduce the readings of the anemometers to the standard level of 10m. In 1977 a commission (North Sea Meteorological Panel) decided upon the reduction factor and all the countries neighboring the North Sea, used that reduction factor in their conversion from observed wind speed at instrument height to a wind speed at 10m.

During that same time several ships were equipped with anemometers as well. Some of these ships were very large and had their anemometers mounted at a great height. The discussion on the value of the recorded wind speeds, with poorly exposed anemometers at a great height on a moving ship, was continuing. Nevertheless no one drew a decisive conclusion.

Generally the use of the ship’s wind speed can be split into three groups:

1. For operational use in models, the reported ship’s wind speed is generally assumed to be reduced to 10m. At present there is no way to find the height of the anemometer and apply a correction accordingly, in real time.
2. For climatological use, the height can be looked up (e.g. in WMO Pub. 47) and a reduction factor can be applied afterwards. There is, however, no way to see whether the observer has applied a correction already or not.
3. Within the VOSCLIM project the use of extra meta-information in the operational SHIP code is not realized. The height of the anemometer is still not available in real time, but still only through WMO Pub. 47. IMMT‑2 for VOSCLIM ships contains, apart from a corrected wind speed and direction, also the relative wind speed and wind direction as well as ships’ course and speed (in TurboWin 2.12 – 3.0). A Data Assembly Centre does the real-time collection of SHIP reports that come from VOSCLIM ships. There the observation is linked to the metadata that is available (WMO Pub. 47 and VOSCLIM recruiting/survey forms).

Starting with TurboWin Release 2.12 a height correction for all types of observing platforms was introduced in the program, to comply with the WMO regulations. This was introduced together with the implementation of the VOSCLIM extension. This, however, stirred up several people. One group favored the new approach; the other group was rather attached to the old style, in which no correction was applied for ships, but only on fixed platforms.

A discussion was held with several experts and finally it was decided that the new release of TurboWin, which will be Release 3.5, should correct the wind speed to 10m for fixed platforms only. Bearing the whole discussion in mind, it is important to remember the following:

• It is unknown whether manufacturers of anemometers use a reduction to 10m in their software. If they do, their methods are not known;

• Wind speeds were corrected to 10m by TurboWin (for ships) in the Releases 2.12 – 3.0; the method is well described and documented;
• On fixed platforms, the measured wind speed is reduced by TurboWin to 10m before transmission of the observation;
• Starting with TurboWin Release 3.5 wind speeds reported by ships, are as they are. No correction for anemometer height is applied;
• In Release 3.5, for VOSCLIM ships, the observers report uncorrected relative wind speed and direction. TurboWin applies a correction for course and speed, not for height, before the message is transmitted. Thus VOSCLIM ships reports (SHIP) contain the true wind speed and wind direction at sensor height. No height of the sensor is transmitted. The VOSCLIM IMMT‑2 record contain both relative wind speed and wind direction at sensor height and the corrected (for course and speed) wind speed and direction at sensor height. No height of the anemometer is available from the IMMT‑2 record;
• Some VOSCLIM ships, however, do not use anemometers to report their wind speeds. These ships use the standard visual estimation techniques. To comply with the IMMT‑2 standard, TurboWin calculates the related relative wind speed for these estimates.
• In Release 3.5, for non-VOSCLIM ships, the observers report uncorrected relative wind speed and direction at sensor height or estimated wind speed and wind direction. TurboWin applies a correction for course and speed, in case of a reported relative wind, not for height, before the message is transmitted. If the observer uses estimated wind speed and wind direction, the wind speed that is transmitted is the wind speed at a (virtual) height of 10m. For this type of ships, the IMMT‑2 records do not contain the reported relative wind speed and wind direction;
• In the “Wind Help File” and on the “Wind Input Screen” TurboWin 3.5 will explicitly show that no correction for height will be applied for ship that report measured wind speeds;
• Being aware of the fact that no alterations to the SHIP code will be approved, the development of the BUFR code (for marine observing platforms) must be followed with care. BUFR promises to hold many nice features, but if the developers come from another space or time this new code will be haunting us, or our – future – colleagues.

Because this matter starts to get confused, I tried to capture the most important features in the following table. The table is set up for wind speed observations on ships (i.e. not wind direction and not on fixed sea platforms) together with TurboWin.

Summary of handling of the wind speed (on board ships) by TurboWin
	Older TurboWin	TurboWin 2.12 – 3.0	TurboWin 3.5
VOSCLIM compliant (IMMT-2)	No	Yes	Yes
SHIP report contains wind speed, “reduced” to 10m	Unknown; only estimated wind speeds are assumed to be given the 10m wind speed	Measured: yes Estimated: yes (implied)	Measured: no Estimated: yes (implied)
IMMT record contains wind speed, reduced to 10m	Unknown; only estimated wind speeds are assumed to be given the 10m wind speed	Yes; VOSCLIM also contains (not reduced) relative wind speed	No
SHIP report contains relative wind speed	No	No	No
IMMT record contains relative wind speed	No	VOSCLIM ship: yes Others: no	VOSCLIM ship: yes Others: no
Relative wind speed corrected for ship speed and course	N.A.	Yes	Yes
Estimated wind speed converted to relative wind speed	No	Wind speeds that are estimated on VOSCLIM ships are converted to relative wind speeds for use in the IMMT record	Wind speeds that are estimated on VOSCLIM ships are converted to relative wind speeds for use in the IMMT record
Anemometer height available in IMMT records or SHIP report?	SHIP: no IMMT: no Pub 47: yes	SHIP: no IMMT: no Pub 47: yes (VOSCLIM recruiting/survey forms contain more info on location of anemometer)	SHIP: no IMMT: no Pub 47: yes (VOSCLIM recruiting/survey forms contain more info on location of anemometer)