EP1828812A1

EP1828812A1 - A weather measurement device for determining the falling speed of hydrometers

Info

Publication number: EP1828812A1
Application number: EP05822873A
Authority: EP
Inventors: Richard Ellis; Jeremy Wolfson School Mech. & Manufac. COUPLAND
Original assignee: Campbell Scientific Ltd
Current assignee: Campbell Scientific Ltd
Priority date: 2004-12-24
Filing date: 2005-12-23
Publication date: 2007-09-05
Also published as: WO2006067513A1; GB2422193A; GB2422193B; GB0428333D0

Abstract

A weather measurement device for determining the speed in a first direction of hydrometeors, the device comprising: a light source for projecting in a second direction, substantially orthogonal to the first direction, a plurality of light sheets each of which has a separation in the first direction with respect to an adjacent light sheets; a light detector for detecting light scattered from a measurement volume occupied by the plurality of light sheets; and a memory for recording the output of the detector whereby sequential scattering events occurring when a hydrometeor traverses the measurement volume in the first direction are recorded in the memory for calculating the speed, in the first direction, of the hydrometeor.

Description

TITLE

A WEATHER MEASUREMENT DEVICE FOR DETERMINING THE FALLING SPEED OF HYDROMETERS

5

FIELD OF THE INVENTION

Embodiments of the invention relate to devices and methods for determining the speed in a first direction of hydrometeors or moving objects. Some 10 embodiments also relate to the subsequent use of the speed determination in classifying precipitation.

BACKGROUND TO THE INVENTION

15 Present weather, as defined by the World Meteorological Organization (WMO) is a description of the weather phenomena present at the time of observation.

Present weather impacts human activity, particularly transportation and can also be used for weather forecasting. Currently human observation is the only 20 method that has the capability to distinguish between all the different forms of weather. Automated present weather systems are, however, much more cost effective and can be deployed in remote and hostile locations where human observations are not practicable.

25 Many different techniques are currently used to measure the speed and size of the hydrometeors found in precipitation.

Doppler radar instruments are also used for present weather sensing. Falling hydrometeors reflect transmitted electromagnetic energy with Doppler shifted 30 frequencies. The shift in frequency is dependent upon the speed of the hydrometeor towards/away from the receiver. The mode frequency of the Doppler power spectrum is used to determine the precipitation type and the total power determines the precipitation rate and intensity. Some present weather sensors can measure visibility and detect precipitation type. Visibility is typically measured using an optical visibility sensor e.g. scattering of an IR beam. The devices also include an analogue capacitive surface sensor and a temperature sensor. The outputs of the sensors are used to identify precipitation type. Although these sensors are good at detecting rain and snow, the discrimination of drizzle, light rain and light snow is less good as described in "WMO lntercomparison of Present Weather Sensors/Systems (1998), WMO/TD No. 887".

Other sensors use direct attenuation to determine the size and speed of hydrometeors. A photo-detector is placed in line of sight with a light source. As a hydrometeor traverses the light beam, the output of the photodiode varies in proportion to the incident light intensity. The output of the photodiode is a pulse with an amplitude dependent on the effective diameter of the hydrometeor and a duration depending on the effective diameter and the hydrometeor velocity.

A 2D video distrometer strongly illuminates falling hydrometeors at two separate positions and captures the silhouette of the falling hydrometeors using a line scan video camera. Complex software can approximate the horizontal velocity of the hydrometeors and determine hydrometeor size, shape, fall velocity etc.

It would be desirable to provide an alterative weather sensor for measuring hydrometeors within precipitation and, in particular, the fall speed of the hydrometeors.

The innovation that is realised in this application is a method similar in principle to Phase Doppler Anemometry (PDA) that has not previously been used in a present weather sensor.

Phase Doppler Anemometry (PDA) is an optical technique that is typically used to measure the speed and size of particles within man made fluid flows. For example, it is used in automotive aerodynamic analysis and fuel nozzle spray analysis. In PDA the output of a laser beam is split into two components of equal intensity. The beams are redirected causing them to cross at a point where they are focussed. The volume where they cross defines the measurement volume. Interference of the light beams in the measurement volume creates a set of equally spaced fringes i.e. bright and dark bands created by a difference of phase of the interfering light waves. A first light detector, positioned so it is not in line of sight of a light beam, measures the light scattered from a particle as it traverses the fringes. The speed S of the particle can be calculated from the period T between the peaks in the detected scattered light and a knowledge of the distance D between the bright bands in the fringes i.e. S= DfT.

A second light detector is used to measure particle size. The second light detector, positioned so it is not in line of sight of a light beam, also measures the light scattered from the particle as it traverses the fringes. When a particle traverses a fringe the scattered light is detected at the first and second detectors with a different phase. The difference in time is dependent upon the size of the particle.

Commercial phase Doppler instrumentation can be used to characterise the particles in the range 0.1 - 100 μm. In order to apply the method to hydrometeors significant design changes must be made.

BRIEF DESCRIPTION OF THE INVENTION

The inventors have realised that the separation of the fringes needs to be increased to enable PDA to be used in a weather sensor for the measurement of the speed of hydrometeors within precipitation. However, the separation of the fringes is constrained by its relationship to the wavelength of the light used which itself must lie within certain boundaries.

According to one aspect of the invention there is provided a number of different embodiments as defined in the appended claims. The inventors have realised that by creating light sheets via direct projection as opposed to fringes via interference, it is possible to control the separation of the light sheets so that they are suitable for use in detecting falling hydrometeors within precipitation.

Embodiments of the invention have several advantages including accurate determination of size and velocity, they are inexpensive compared to video distrometers, and are capable of measuring a wide range of present weather variables which allows them to be a viable replacement to human observation.

BRIEF DESCRIPTION OF THE DRAWINGS

For a better understanding of embodiments of the present invention and to understand how they may be practised, reference will now be made by way of example only to the accompanying drawings in which:

Fig. 1 illustrates an automated weather measurement device;

Fig. 2 schematically illustrates one suitable arrangement that can be used as a light source; Fig. 3 schematically illustrates one suitable arrangement that can be used as a light detector;

Fig. 4 schematically illustrates some of the functional components of the weather measurement device;

Figs 5A and 5B are flow diagrams illustrating the operation of the weather measurement device.

DETAILED DESCRIPTION OF EMBODIMENTS OF THE INVENTION

Fig. 1 illustrates an automated weather measurement device 10 that measures precipitation within a measurement volume 12. In particular it measures the speed of hydrometeors 1 falling through the measurement volume 12 in a first vertical direction 14. The size of the falling hydrometers are typically in the range 0.1 to 30 mm diameter which corresponds to hydrometeors found within precipitation. The measurements of size and speed can be used with measurements of temperature and humidity to classify the precipitation type.

The device 10 uses the basic principle of PDA in that a regular series of light sheets 14 are observed in a measurement volume 12, there is detection of scattering events at two different angles as a hydrometeor 1 falls through the measurement volume 12 and there is frequency and phase analysis of the detected scattering events to determine the speed and size values for the hydrometeor 1. The area of the measurement volume in the plane of the light sheets (i.e. the volume footprint) is of the order 10 cm² to 60 cm², to receive enough particles in the range of precipitation events to make statistical calculations and be sensitive enough without over saturating the system with too many particles. Using other processing techniques it may be possible to increase the measurement volume. The sheet separation is typically between 0.5 mm and 1 mm to cover the full range of particle sizes and speeds but sheet separations outside this range may be used.

The weather measurement device 10 comprises: a light source 2; a first light detector 20 for detecting light scattered from a measurement volume 12; a second light detector 30 for detecting light scattered from the measurement volume 12; and a control unit 6 comprising a processor 40, a memory 42 and an output interface 44.

The light source 2 projects in a second horizontal direction 16, substantially orthogonal to the first vertical direction 14, a plurality of light sheets 4 that have a constant separation in the first direction between adjacent light sheets. The constant separation between the light sheets is of the order of 0.5 mm to 1 mm. There are two or more light sheets 4 and each sheet forms a horizontal plane of light that is substantially orthogonal to the first direction within the measurement volume 12.

The first and second light detectors 20, 30 are positioned in a forward scattering configuration and are directed towards the measurement volume 12. This defines a measurement volume of the order 1 cm³ to 60cm³ that is approximately ellipse and gives good signal strength from all types of precipitation- drizzle, rain, snow, sleet, hail etc. The first light detector 20 lies in a horizontal plane and is oriented at a 20 degree deflection angle to the second direction 16. It collects the light 5 scattered from the measurement volume 12 by falling hydrometeors 1 The second light detector 30 lies in a vertical plane and is oriented at a 20 degree deflection angle to the second direction 16. It collects the light 7 scattered from the measurement volume 12 by falling hydrometeors 1.

The light source 2 and light detectors 20, 30 are mounted in respective head units that are mounted in fixed positions using a rigid mast 3. The mast 3 is used to route cables between the control unit and head units and ensures protection from the elements. The head units and mast 3 are joined using high strength adhesive to avoid manufacturing errors inherent in welding, due to heat deformation. Head units may have heated hoods and a letterbox shaped baffle.

Fig. 2 schematically illustrates one suitable arrangement that can be used as the light source 2. A single laser diode 202 projects light through a cylindrical lens 204 onto a grating 206. The grating 206 is opaque with a series of horizontal, parallel transparent slit apertures that are equally separated in the vertical direction 14. The grating 206 only allows the passage of the light beam through the transparent areas forming a number of separate light sources. The slit apertures are periodically spaced and have periodicity in the vertical direction 14. The light sheets pass through a rod lens 208 and doublet lens 210, which form the light sheets into parallel, horizontal light planes 16 that are orthogonal to the vertical direction 14. These light planes 16 are output by the light source 2 and form a pattern that has periodicity in the vertical direction 14 and negligible distortion in the measurement volume 12. The separation between the light sheets in the measurement volume 12 is between 0.5 mm and 1 mm.

The light source 2 is thus arranged to simultaneously project parallel light sheets 2 within the measurement volume 12 that have a constant separation of between 0.5 mm and 1 mm. There are other arrangements of light source 2 that can be used to form parallel light sheets 2 within the measurement volume 12 that have a constant separation of between 0.5 mm and 1 mm, for example the light sheets could be projected directly from the laser 202 or other arrangements of optics than those disclosed in Fig. 2 can be used.

It is also possible to use a light source 2 that sequentially projects parallel light planes 16 within the measurement volume 12 that have a constant separation of between 0.5 mm and 1 mm. This is achieved by scanning a single light plane output by a laser diode 202 back and forth over the measurement volume 12 at high frequency.

Fig. 3 schematically illustrates one suitable arrangement that can be used as a light detector 20, 30. The scattered light 5, 7 is focused by a plano-convex lens 306, through a filter 304 onto a photodiode 302.

Fig. 4 schematically illustrates some of the functional components of the weather measurement device 10.

The output of the light source 2 is modulated using a high frequency signal produced by generator 401 as an input to the laser driver 402 that controls the laser light source 2. The modulation of the light source 2 is at high frequencies typically in the range of tens to hundreds of kHz.

Sequential scattering events occur when a hydrometeor traverses the measurement volume 12 in the first direction 14. A scattering event occurs each time the hydrometeor 1 traverses a light sheet 16.

The scattered light 5 is captured by the first light detector 20 which produces an output voltage value that is proportional to the incident light intensity. This signal is demodulated and stored in memory 42. The scattered light 5 is scattered by the second light detector 30 which produces an output voltage value that is proportional to the incident light intensity. This signal is demodulated and stored in memory 42.

The processor 40 processes the recorded outputs and the output interface 44 is used to output the results of the processing. The output interface 44 may be a galvanic interface or a wireless interface such as a radio transceiver.

The modulation/demodulation used in the weather measurement device is used to remove noise from the output of the light detectors 20, 30 arising from other light sources than the scattered light that vary more slowly that the frequency of modulation. When the laser is switched off, the light detector detects a background value and when the laser is switched on the light detector detects the scattered light and the background value. The output of the light detector can consequently be corrected to remove the background so that it includes only a value associated with the scattered light 5, 7.

The weather measurement device 10 may additionally house sensors for measuring ambient parameters such as a temperature sensor 411 , a humidity sensor 412 and a wetness sensor 413. The outputs of the sensors are also stored in the memory 42.

The storage of recorded data retains the time information concerning the data. The memory 42 buffers the output of the sensors for future processing by the processor 40. A first in first out (FIFO) memory is suitable.

The weather measurement device 10 may additionally have a mechanism for compensating for a reduction in the output power of the light source 2 arising from dirt accumulating on the external lens 210 of the light source 2. A light emitting diode (LED) 420 is housed within the light source and it directs lights onto the external lens. A photodiode 422 also housed within the light source 2 is used to detect the amount of light reflected back by the lens. This can be used to detect dirt on the outer lens 306. The output of the photodiode 422 is also stored in the memory 42 for recalibrating the outputs of the first and second light detectors 20, 30 as the output power of the light source 2 decreases because of the accumulation of dirt.

The operation of the processor 40 will now be described with reference to Figs 5A and 5B.

At step 50 a visibility calculation is performed. The processor 40 detects the dc offset in the output data from a light detector and uses this value to quantify the visibility. The DC offset in the output of the detector is due to widespread light scatter and a higher value indicates lower visibility.

Then at step 52 at system check is performed. For example, the output from the photodiode 422 may be checked to ensure that the lens of the light source 2 is sufficiently clear for use.

At step 54, the output data from the light detectors 20, 30, the ambient sensors 411 , 412, 413 and the photodiode 422 are buffered in the memory 42.

At step 56, the analysis of the buffered data begins. Initially the analysis is in the time domain. The processor analyses the output of the first light detector 20 to identify a portion in time where the output value exceeds a predetermined threshold. Such a significant scattering event is typically associated with a large hydrometeor. If the presence of such a large hydrometeor is detected the process moves through step 58 to step 60. If a significant scattering event is not present in the buffered data the process moves to step 68.

At step 60 the data from the first light detector is analysed in the region of the identified significant scattering event to identify the beginning and end of the series of significant scattering events associated with the large hydrometer traversing the parallel light planes 16. The scattering event data representing the series of significant scattering events is then analysed at step 62 to determine the large hydrometeors velocity. A fast Fourier transform is taken of the scattering event data which identifies the periodicity P in Hz between significant scattering events. Each of these significant scattering events is associated with one of the light planes 16 within the measurement volume. If the light planes have a separation D in metres, then the speed in metres per second of the large hydrometeor in the first direction 14 is given by D ^* P.

Steps 60 and 62 may also be carried out for the data output by the second light detector 30 to obtain an alternative value for the speed of the large hydrometeor.

The scattering event data for the first and second detectors representing the same series of significant scattering events are then analysed at step 64 to determine the size of the large hydrometeor that caused the series by traversing the light planes 16. A fast Fourier transform is taken for the scattering event data for the first detector and a fast Fourier transform is taken for the scattering event data for the second detector. The cross correlation is then calculated using the FFT data to obtain a value for the phase offset between the series of scattering events for the first and second detectors. This phase offset has a known and fixed relationship with the size of the hydrometeor.

At step 66, the scattering event data for the first and second detectors representing the same series of significant scattering events are removed from the stored output data, so that further processing of the output data will not re-use the already processed data.

The process then returns to step 56.

At step 68, the analysis of the buffered output data continues with analysis in the frequency domain. The processor performs an FFT on the output of the first light detector 20 and identifies a portion in frequency where the output value exceeds a predetermined threshold. Such a scattering event is typically associated with a smaller hydrometeor. If the presence of such a hydrometeor is detected the process moves through step 70 to step 72. If a scattering event is not present in the buffered data the process moves to step 80.

At step 72 the output data from the first light detector is analysed to identify the beginning and end of a series of the scattering events. A FFT scan is used to identify signals in the data that are too small to be identified from the time domain data. The identified peak on the FFT can be used for velocity and size processing or can be reconstructed into the time domain with the pertinent frequency only.

Further processing can then correctly identify the start and end of the data without the noise from other frequencies not associated with the particle falling through the detection volume. The resulting signal can then be used in the previously described way to determine the particle size.

The scattering event data representing the series of scattering events is analysed at step 74 to determine the hydrometeor's velocity. A fast Fourier transform is taken of the scattering event data which identifies the periodicity P in Hz between scattering events within the series. Each of these scattering events is associated with one of the light planes 16 within the measurement volume. If the light planes have a separation D in metres, then the velocity in metres per second of the large hydrometeor is given by D ^* P.

Steps 72 and 74 may also be carried out for the data output by the second light detector 30 to obtain an alternative value for the speed of the large hydrometeor.

The scattering event data for the first and second detectors representing the same series of significant scattering events are then analysed at step 76 to determine the size of the hydrometeor that caused the series by passing through the light planes 16. A fast Fourier transform is taken for the scattering event data for the first detector and a fast Fourier transform is taken for the scattering event data for the second detector. The cross correlation is then calculated using the FFT data to obtain a value for the phase offset between the scattering event data for the first and second detectors. This phase offset has a known and fixed relationship with the size of the hydrometeor.

At step 78, the scattering event data for the first and second detectors representing the same scattering event are removed from the stored output data, so that further processing of the output data will not re-use the already processed data.

The process then returns to step 68.

At step 80, the processor 40 determines whether or not a time constant has been met. If it has, the processor at step 82 classifies the obtained size and velocity values, but if it has not then the process returns to step 54 to capture more data from the light detectors 20, 30.

At step 82, the processor 40 classifies the type of precipitation using the determined speed and size measurements of the hydrometeors.

The relationship of size to speed can be used to differentiate snow from rain. As the size of a rain droplet increases so does its speed. However, as the size of a snow flake increases its speed does not necessarily increase. If the size and speed of each hydrometeor were plotted on a size vs. speed scatter graph more scatter indicates a greater probability that the hydrometeors are snow flakes.

The classification may additionally use measurements of humidity and temperature from the respective sensors 41 1 and 412 to discriminate between precipitation types.

It is possible to determine the most likely precipitation type for a given set of speed, size, temperature and humidity by using fuzzy logic. For each precipitation type, different temperature ranges are assigned a probability value between 0 and 1. For each precipitation type, different humidity ranges are assigned a probability value between 0 and 1. For each precipitation type, different size and speed combinations are assigned a probability value between 0 and 1. For a given hydrometeor of a certain size and speed with certain humidity and temperature, it is possible to determine the probability that the hydrometeor is a certain precipitation type. This value coupled with those from the fuzzy logic matrices of auxiliary instruments gives a more accurate prediction of particle type.

It is also possible to use the signal structure to help define the particle type. Snow tends to produce signals with larger pedestals and smaller peaks due to the complex structure of the particles. This can be used to distinguish particle types through average and peak signal values, ratios of which are assigned fuzzy logic values. For snow, which can give rise to multiple reflections and refractions from the crystalline surfaces, the signal intensity can be used as a further parameter for determining particle type.

For example, if the humidity is less than 40% the likelihood of rain is very small and as the humidity increases the likelihood increases. If the temperature is greater than 0 degrees Celsius the likelihood of snow is small and but as the temperature decreases the likelihood of snow increases.

The processor is operable to classify precipitation as one of, for example, drizzle, freezing drizzle, falling snow, blowing snow, snow grains, rain, freezing rain, ice pellets and hail. Further particle types, graupel and ice crystals for example, may also be distinguished.

In the foregoing description, the processor 40 is described as being co-located with the memory 42, light source 2 and light detectors 20, 30. In other implementations, it would be possible to remove the processor 40. In this case the collated data stored in the memory 40 would be transmitted to a remote site for processing there. It should be appreciated that although the light detectors have been described as having specific angles of deflection to the second direction it is possible to use other angles of deflection. Also although the light detectors illustrated are deflected in orthogonal directions from the second direction 16 this is not essential and the light detectors 20, 30 could lie in the same plane. However, the light detectors are positioned off-axis with respect to the light source 2, so that they are not directly in line of sight with the light source 2.

It would be possible to add a third light detector, which would be deflected at an angle to the second direction 16. The three different pairs of detectors would enable three different size values for a hydrometeor to be calculated.

Although the light source 2 has been described as a laser, this is not essential and any suitable light source may be used. Lasers may be preferred because of their high power and spatial coherence.

Although embodiments of the invention have been described in relation to a weather measurement device 10 that measures the speed and size of hydrometeors within precipitation, the invention has other applications where it is necessary to measure the speed and size of particles that have a size greater than a few millimetres.

Although embodiments of the present invention have been described in the preceding paragraphs with reference to various examples, it should be appreciated that modifications to the examples given can be made without departing from the scope of the invention as claimed.

Whilst endeavouring in the foregoing specification to draw attention to those features of the invention believed to be of particular importance it should be understood that the Applicant claims protection in respect of any patentable feature or combination of features hereinbefore referred to and/or shown in the drawings whether or not particular emphasis has been placed thereon.

I/we claim:

Claims

1. A weather measurement device for determining the speed in a first direction of hydrometeors, the device comprising:

a light source for projecting in a second direction, substantially orthogonal to the first direction, a plurality of light sheets each of which has a separation in the first direction with respect to an adjacent light sheets;

a light detector for detecting light scattered from a measurement volume occupied by the plurality of light sheets; and

a memory for recording the output of the detector whereby sequential scattering events occurring when a hydrometeor traverses the measurement volume in the first direction are recorded in the memory for calculating the speed, in the first direction, of the hydrometeor.

2. A weather measurement device as claimed in claim 1 , further comprising: processing means for processing the output of the light detector to identify sequential scattering events that occur when a hydrometeor traverses the measurement volume and for calculating, using the separation of the light sheets, the speed in the first direction of the hydrometeor.

3. A weather measurement device as claimed in claim 2, further comprising: means for identifying a first portion of the output of the light detector comprising scattering events, associated with a large hydrometeor, by determining when the output of the detector exceeds a threshold; means for analysing the first portion of the output to determine the periodicity of the scattering events associated with the large hydrometeor; and means for using the determined periodicity to determine the speed of the large hydrometeor in the first direction.

4. A weather measurement device as claimed in claim 2 or 3, further comprising: means for modifying the output data to remove a first portion; means for performing frequency analysis of the remaining data to identify a second portion of the output data that records scattering events associated with a second hydrometeor; and means for performing frequency analysis of the second portion of data to determine the periodicity of the scattering events associated with the second hydrometeor and for calculating the speed of the second hydrometeor in the first direction.

5. A weather measurement device as claimed in claim 2, wherein the separation between respective light sheets is the same, the device further comprising means for analysing the output of the first light detector, determining the periodicity of the scattering events and calculating the speed of the hydrometeor in the first direction from the determined periodicity.

6. A weather measurement device as claimed in claim 2, further comprising: means for identifying scattering events, associated with a large hydrometeor, within the output of the light detector by determining when the output of the light detector exceeds a threshold; means for determining the timing between scattering events associated with the large hydrometeor; and means for using the determined timings to determine the speed of the large hydrometeor in the first direction.

7. A weather measurement device as claimed in any preceding claim wherein the light detector is oriented at 20 degrees to the second direction in a forward scattering configuration.

8. A weather measurement device as claimed in any preceding claim, further comprising means for pulsing the light source at a predetermined high frequency.

9. A weather measurement device as claimed in any preceding claim, wherein the light source is arranged to simultaneously project the light sheets within the measurement volume.

10. A weather measurement device as claimed in claim 14, further comprising a single light emitter and a mask having an aperture for each light sheet.

11. A weather measurement device as claimed in any one of claims 1 to 8, wherein the light source is arranged to sequentially project each one of the plurality of light sheets within the measurement volume.

12. A weather measurement device as claimed in claim 11 , wherein the light source produces a single light sheet that is repeatedly scanned across the measurement volume.

13. A weather measurement device as claimed in any preceding claim, further comprising a further, second light detector for detecting light scattered from the measurement volume, wherein the memory is operable to record the output of the second detector thereby recording, when a hydrometeor traverses the measurement volume and causes a scattering event, a difference in time between the detection of the scattering event at the respective detectors that corresponds to the size of the hydrometeor.

14. A weather measurement device as claimed in claim 13, further comprising processing means for calculating the size of the hydrometeor using the difference in time between the detection of scattering events at the respective detectors.

15. A weather measurement device as claimed in claim 14, wherein the processing means cross-correlates the outputs of the light detectors in the frequency domain.

16. A weather measurement device as claimed in claim 13, 14 or 15 further comprising classification means for identifying a current type of precipitation using the speed and size measurements of hydrometeors.

17. A weather measurement device as claimed in claim 16 wherein the classification means additionally uses measurements of humidity and temperature to discriminate between precipitation types.

18. A weather measurement device as claimed in any preceding claim that is operable as a present weather sensor, the device further comprising: means for detecting a dc offset in the output of the light detector; and means for quantifying visibility using the d.c. offset.

19. A weather measurement device as claimed in any preceding claim, wherein the separation between the plurality of light sheets is in the range 0.5 mm to 1 mm.

20. A weather measurement device as claimed in any preceding claim, wherein the measurement volume is in the range 1 cm³ to 60 cm³.

21. A weather measurement device as claimed in any preceding claim, wherein the plurality of light sheets includes two or more light sheets.

22. A weather measurement device as claimed in any preceding claim, wherein the separation between any adjacent pair of light sheets is the same.

23. A weather measurement device as claimed in any preceding claim, wherein each sheet of light is a plane of light that is substantially orthogonal to the first direction within the measurement volume.

24. A method of weather measurement for determining the speed in a first direction and size of hydrometeors within precipitation, the method comprising: projecting in a second direction, substantially orthogonal to the first direction, a plurality of light sheets each of which has a separation in the first direction with respect to an adjacent light sheet; detecting using a light detector at a first position light scattered from a measurement volume occupied by the plurality of light sheets when a hydrometeor traverses the measurement volume; and calculating the speed in a first direction of the hydrometeor using the output of the light detector.

25. A method as claimed in claim 24, wherein the light sheets are projected simultaneously.

26. A method as claimed in claim 24, wherein the light sheets are projected sequentially.

27. A method as claimed in claim 24, 25 or 26, wherein the separation between the light sheets is of the order of 0.5 mm to 1 mm.

28. A method as claimed in any one of claims 24 to 27, wherein the measurement volume is of the order 1 cm³ to 60 cm³.

29. A method as claimed in any one of claims 24 to 28, wherein each light sheet is a plane of light substantially orthogonal to the first direction in the measurement volume and the planes of light are equally separated.

30. A method as claimed in any one of claims 24 to 29, wherein calculating the speed in a first direction of the hydrometeor involves analysing the output of the light detector to determine the periodicity of the scattering events caused by a hydrometeor moving in the first direction through the measurement volume and using the periodicity to calculate the speed of the hydrometeor in the first direction.

31. A method as claimed in any one of claims 24 to 30, wherein calculating the speed in a first direction of the hydrometeor involves: using a threshold to identify a first portion of the output from the light detector that records scattering events associated with a large hydrometeor; analysing the first portion of data to determine the periodicity of the scattering events associated with the large hydrometeor; and using the periodicity to calculate the speed of the large hydrometeor in the first direction.

32. A method as claimed in claim 31 , wherein calculating the speed in a first direction of a small hydrometeor involves modifying the output from the light detector to remove the first portion; performing frequency analysis of the remaining output to identify a second portion of the output that records scattering events associated with a small hydrometeor; performing frequency analysis of the second portion to determine the periodicity of the scattering events associated with the small hydrometeor; and using the periodicity to calculate the speed of the small hydrometeor in the first direction.

33. A method as claimed in any one of claims 24 to 32, further comprising pulsing the light sheets at a predetermined high frequency.

34. A method as claimed in any one of claims 24 to 33, further comprising detecting, using a second light detector at a second position, light scattered from the measurement volume occupied by the plurality of light sheets when a hydrometeor traverses the measurement volume; determining a difference in time between the detection of a scattering event at the respective first and second positions; and using the difference in time to calculate the size of the hydrometeor.

35. A method as claimed in claim 34, further comprising: detecting, using a third light detector at a third position, light scattered from the measurement volume occupied by the plurality of light sheets when a hydrometeor traverses the measurement volume; determining differences in time between the detection of a scattering event at the respective first, second and third positions; and using the differences in time to calculate the size of the hydrometeor.

36. A method as claimed in claim 34 or 35, wherein the determined speed and sizes of hydrometeors are used to identify a precipitation type within the measurement volume.

37. A method as claimed in claim 36, wherein the identification of a precipitation type within the measurement volume additionally uses measurements of humidity and temperature.

38. A method as claimed in any one of claims 24 to 36 further comprising detecting a dc offset in the detected scattered light and using the detected dc offset to determined a quantitative measure of visibility.

39. A use of phase Doppler anemometry for determining the speed in a first direction and size of hydrometeors within meteorological precipitation wherein the measurement volume is defined using a light source that projects in a second direction, substantially orthogonal to the first direction, a plurality of light sheets each of which has a fixed separation in the first direction with respect to an adjacent light sheet and is substantially orthogonal to the first direction.

40. A weather measurement device comprising: a light source for projecting a plurality of light sheets each of which has a separation in the first direction with respect to an adjacent light sheet; a first light detector for detecting light scattered from a measurement volume occupied by the plurality of light sheets; a second light detector for detecting light scattered from the measurement volume; means for processing the outputs of the first and second light detectors to identify a type of precipitation within the measurement volume.

41. A measurement device for determining the speed in a first direction of objects, the device comprising: a light source for projecting in a second direction, substantially orthogonal to the first direction, a plurality of light sheets each of which has a separation in the first direction with respect to an adjacent light sheet; a light detector for detecting light scattered from a measurement volume occupied by the plurality of light sheets; and a memory for recording the output of the light detector whereby sequential scattering events occurring when an object traverses the measurement volume in the first direction are recorded in the memory for calculating the speed, in the first direction, of the object.

42. A measurement device as claimed in claim 41 for measuring the speed of objects having a maximum dimension greater than 0.1 mm.

43. A weather measurement device substantially as hereinbefore described and/or as shown in the accompanying drawings.

44. Any novel subject matter or combination including novel subject matter disclosed, whether or not within the scope of or relating to the same invention as the preceding claims.