1. © Crown copyright Met Office Cardington Turbulence Probes James McGregor (OBR Cardington) OBR Conference, 12 th December 2012

2. © Crown copyright Met Office Brief History 1940s : Meteorological measurements using a tethered balloon system first made at Cardington BALTHUM ascents Late 1960s : First attempt at developing a turbulence probe Damped pendulum system Mid 1980s : Development of “New” (now old!) turbulence probe Uses inclinometers and magnetometers Algorithms for correcting for cable motion Platform : 24,000 cubic ft kite balloon, stainless steel cable Probe weight : 10kg Logging and processing on VAX machines 2000 : Retirement of large balloon system

3. © Crown copyright Met Office Old tethered balloon system

4. © Crown copyright Met Office New turbulence probe 2002 : First trials carried out on prototype lightweight probe Modelled on old probe Small tethered balloon with lightweight polyethylene fibre cable Do the motion correction algorithms work effectively on new platform? 2010 : Total six probes constructed 2011-2012 : Updated software Logging software migrated onto windows PC Processing software rewritten in IDL

5. © Crown copyright Met Office New Turbulence Probe 3 propeller anemometers. Starting speed ~0.5 m/s Inclinometers for pitch and roll 3-axis fluxgate magnetometer for orientation Temperature/humidity sensors in aspirated screen Fast response temperature (thin-wire PRT) Pressure sensor Tail fin keeps probe pointing into wind Li–ion battery: duration up to 8 hours Total weight 1.2kg

6. © Crown copyright Met Office Tethered Balloon 1800 Cuft capacity kite balloon (~30ft long) Trailer mounted hydraulic winch, driven by a small Honda petrol engine. 2km (>6000ft) tether cable 25kg static lift at surface 7kg net lift approx at 6000ft Max wind aloft for flying ~30kts (15 kts at surface for balloon handling) Tensiometer and odometer (measures length of cable paid out) incorporated

7. © Crown copyright Met Office Applications Profiling Fog studies Stable boundary layer Level runs Turbulence below Sc clouds Often flown in conjunction with other instruments:- Aerosol measurements Cloud droplet probe

8. © Crown copyright Met Office Probe Logging 400 MHz (Met band) data telemetry system Probe data sampled at ~20 Hz and transmitted using “Manchester” encoded signal Ground station consists of a series receivers and Manchester decoder boards New logging software recently developed on windows PC Real time data displays – important feature of system

9. © Crown copyright Met Office Probe logging program display

10. © Crown copyright Met Office Real time graphical display Graphical displays written in IDL Displays updated in real time Choice of variables to plot (T, TD, RH, Q, UTOT, DIRN etc) Profile plots or time-series

11. © Crown copyright Met Office Probe Calibrations Temperature/humidity cals in Cardington environmental chamber Custom built rigs for magnetometer and inclinometer calibrations Pressure calibration rig Anemometers calibrated in wind tunnel. Propeller cosine response and step response tests also carried out in wind tunnel

12. © Crown copyright Met Office Propeller Length constant

13. © Crown copyright Met Office Propeller cosine response Propeller stalls at wind angle of ~82 degrees Non-symmetric response around 90 deg Enhanced output in reverse direction? Bump between 165 and 180 due to shielding in wind tunnel Likely to be errors in light/variable winds

14. © Crown copyright Met Office Probe Processing 1.Calibrate sensors 2.Transform measured winds into orthogonal components relative to the probe body 3.(Apply frequency response corrections to propeller anemometers) 4.Calculate cable orientation using inclinometers and magnetometer 5.Time filter orientation to remove spurious data caused by inclinometer accelerations –cable orientation assumed to vary slowly with time. 6.Orientation of probe around cable axis derived from magnetometer alone – not subject to acceleration affects 7.Derive U,V and W in ground-based reference frame 8.Filter out probe motion – software attempts to remove two modes of cable motion from the data …

15. © Crown copyright Met Office Cable motion 1 – catenary motion Low frequency translational motion caused by drift of balloon Period of motion of the order of a minute Use filtered cable orientations and probe heights to construct approximate shape of cable catenary Differentiate probe positions in northerly, westerly and vertical directions to obtain probe velocities Add probe velocities to measured U, V and W wind components. Accuracy of algorithm is improved with multiple probes on the cable Correction may not be as effective using a single probe at a high altitude

16. © Crown copyright Met Office Cable motion 2 – Harmonic motion Harmonic oscillations of cable – period of a few seconds Causes over estimate of turbulence variables Use inclinometers as accelerometers at frequencies over which cable tilt is not expected to alter significantly Measured accelerations are integrated to give probe velocities Frequency range of oscillations must be well separated from catenary motion for corrections to work

17. © Crown copyright Met Office Data Analysis Plot velocity power spectra obtained from probe (U,V and W) as a function of frequency Typical boundary layer turbulence spectrum has energy created at longer wavelengths and decaying through an inertial sub- range following a 5/3 power law Extra peaks in spectra caused by probe motions Further peak in V spectra may be apparent due to limitations in propeller frequency response

18. © Crown copyright Met Office Velocity Power Spectrum Density  Probe height: ~400m  Wind speed ~3m/s  No motion corrections applied  Peaks mainly show up in the V component of turbulent energy

19. © Crown copyright Met Office Velocity Power Spectrum Density  Probe height: ~400m  Wind speed ~3m/s  Catenary motion correction applied

20. © Crown copyright Met Office Velocity Power Spectrum Density  Probe height: ~400m  Wind speed ~3m/s  Catenary motion correction applied  Accelerometer corrections applied

21. © Crown copyright Met Office Velocity Power Spectrum Density  Probe height: ~400m  Wind speed ~3m/s  Catenary motion correction applied  Accelerometer corrections applied  Inertial sub-range decay line (-5/3 power law)  Slight residual peak in V component  Sharp drop off at high frequencies due to limited frequency response of propellers

22. © Crown copyright Met Office Velocity Power Spectrum Density  Probe height: ~400m  Wind speed ~3m/s  Catenary motion correction applied  Accelerometer corrections applied  Inertial subrange decay line (-5/3 power law)  Propeller response correction applied

23. © Crown copyright Met Office Future Work Software:- Further testing of balloon motion correction algorithms Electronics:- Upgrade outdated analogue probe circuitry – timing issues Upgrade ground station receivers and decoder boards Sensors:- Sort out problem with humidity sensors (staying saturated after exiting from cloud) New pressure sensor? Consider alternative propeller anemometers? Next generation probe – replace propellers with sonic anemometer?

24. © Crown copyright Met Office Questions and answers