Evaluation of SVPBW drifters thanks to deployments near
Evaluation of SVP-BW drifters thanks to deployments near moored buoys By Pierre Blouch Presentation : Jean Rolland DBCP-18 workshop - Martinique 14 -15 October 2002
Background ü About 280 SVP type drifters reporting wind data have been deployed since 1997 : - 165 from Metocean (WOCE WSD/BP) - 115 from Pacific Gyre (Minimet) ü All of them are using the WOTAN technique to estimate the wind speed ü Common name at DBCP : SVP-BW drifters DBCP-18 workshop – 14 -15 Oct. 2002
What’s the WOTAN technique ? ü WOTAN means Wind Observation Through Ambient Noise ü WOTAN was also the Sun God in the Germanic Mythology ü The WOTAN technique uses the relationship between : - the energy of the submarine ambient noise - and the wind speed blowing above the sea surface DBCP-18 workshop – 14 -15 Oct. 2002
Generic underwater sound spectra according to Nystuen & Selsor (1997) 100 Spectral level (d. B re 1 m. Pa²/Hz) hail 20 80 Wind m/s Heavy rain 12 5 60 Light rain, no wind 2 Light rain, 3 m/s wind 40 snow 2 k. Hz 20 0. 1 1 8 k. Hz 10 Frequency (k. Hz) DBCP-18 workshop – 14 -15 Oct. 2002 100
Aims of the present study ü To compare wind observations provided by Metocean SVP-BW drifters with those reported by moored buoys ü To check the parameters and the algorithm used by Metocean SVP-BW drifters to provide wind speed data ü To adjust these parameters and to improve the algorithm if necessary and possible DBCP-18 workshop – 14 -15 Oct. 2002
What’s a Metocean SVP-BW drifter ? Hydrophone Fixed vane Barometer port low e e sb c rt e urfa e m ea s 0 1 es th Drogue Drifter hull DBCP-18 workshop – 14 -15 Oct. 2002
Sound spectra from a Metocean SVP-BW drifter ~ 17 m/s 8 k. Hz ~ 4 m/s 2 k. Hz DBCP-18 workshop – 14 -15 Oct. 2002
Trajectory of drifter 32783 28 days - but drogue lost after 3 days DBCP-18 workshop – 14 -15 Oct. 2002
78 days On October 3 rd, 2002 drifter 32792 was close to moored buoy 44140 Trajectories of drifters 32791 & 32792 6 days 44 days
Comparisons for wind direction DBCP-18 workshop – 14 -15 Oct. 2002
Algorithm used to provide wind speed estimates ü Wind speed estimate at 10 meters height : U 10 = a * 10 SL(f)/20 + b where SL(f) is the sound intensity at frequency f expressed in d. B relative to 1 m. Pa 2/Hz a and b are two empirical coefficients depending on the frequency ü Metocean SVP-BW drifters compute wind speed estimates for 2 and 8 k. Hz. Coefficients for 2 k. Hz are presently wrong. DBCP-18 workshop – 14 -15 Oct. 2002
Comparisons for wind speed Wrong values declared as good Good values declared as wrong DBCP-18 workshop – 14 -15 Oct. 2002
Relationship between wind speed and sound level at 2 k. Hz 17. 5 m/s DBCP-18 workshop – 14 -15 Oct. 2002 Slope change
Relationship between wind speed and sound level at 4 k. Hz 17. 5 m/s DBCP-18 workshop – 14 -15 Oct. 2002 Slope change for all frequencies
Linear relationship between wind speed and Po Best correlation 2 k. Hz DBCP-18 workshop – 14 -15 Oct. 2002 6 k. Hz
4 k. Hz DBCP-18 workshop – 14 -15 Oct. 2002 Wind speed vs frequency and Po (adjustment)
Sound spectra (adjustment) Best correlation 6 k. Hz 2 k. Hz DBCP-18 workshop – 14 -15 Oct. 2002
Algorithm proposed to provide better wind speed estimates ü Wind speed estimate at 10 meters height is the average of estimates computed at 2, 3. 15, 4, 5 and 6 k. Hz ü Wind speed estimate is flagged if the standard deviation of estimates computed between 1 and 8 k. Hz is higher than 2. 5 m/s DBCP-18 workshop – 14 -15 Oct. 2002
Proposed wind speed computation Standard deviation greater than 2. 5 m/s DBCP-18 workshop – 14 -15 Oct. 2002
Present wind speed computation (reminder) DBCP-18 workshop – 14 -15 Oct. 2002
Conclusions (1) ü Wind speed estimates presently computed from 8 k. Hz are inaccurate. This frequency isn’t the most suitable for the searched goal. ü Calibration values presently used for 2 k. Hz are incorrect ü The best correlation between sound energy and wind speed is obtained between 2 and 6 k. Hz DBCP-18 workshop – 14 -15 Oct. 2002
Conclusions (2) ü A slope change occurs at about 17. 5 m/s in the relationship between sound energy and wind speed. This could be due to breaking waves. ü The quality of wind speed estimates is significantly improved by using an average between 2 and 6 k. Hz ü The standard deviation of wind speed estimates computed between 1 and 8 k. Hz can be used to reject dubious values from GTS transmission. DBCP-18 workshop – 14 -15 Oct. 2002
Conclusions (3) ü Calibration values found here are valid for present Metocean WOTAN drifters only ü Software modifications will be implemented on our next Metocean SVP-BW drifters ü Evaluation should continue, mainly on strong winds (i. e. higher than 20 m/s) thanks to drifters providing sound spectra DBCP-18 workshop – 14 -15 Oct. 2002
Acknowledgements to ü UK Met Office for deployment near K 3 moored buoy ü Environment Canada for the two other deployments ü Metocean for the software modifications inside the drifters ü Any body who would help us to continue this evaluation in rough seas by providing sound spectra from some drifters deployed near moored buoys ü … DBCP-18 workshop – 14 -15 Oct. 2002
- Slides: 25