Titre de la présentation - CONFIDENTIAL 26/05/2014 - 1
Combined wind & aerosol/cloud measurements with coherent Doppler LIDARs
for operational networks
Ludovic Thobois, Philippe ROYER,
Jean Pierre Cariou
Titre de la présentation - CONFIDENTIAL 26/05/2014 - 2 03/11/2015 - 2 TOPROF WC5 Meeting, Toulouse
Key specifications of Doppler LIDARs
Figure of Merit: Allows to compare different lidars or different configurations. It allows to classify LIDAR sensitivities, independently of atmospheric parameters. WINDCUBE7v2 ~5.6E-12 WINDCUBE200S ~3.39E-10
Range resolution Maximum operational range
Titre de la présentation - CONFIDENTIAL 26/05/2014 - 3 03/11/2015 - 3 TOPROF WC5 Meeting, Toulouse
Key specifications of Doppler LIDARs
Range resolution: Is defined as the Full Width Half Maximum (FWHM) of the averaging kernel For pulses and weighted range gates
Squared Gaussian
WINDCUBE7v2 ~25m WINDCUBE200S ~50-150m
Maximum operation range Maximum distance where a confident wind speed can be derived from the lidar signal It must be defined for a given data availability ie 50% or 80%
Rang
e
Cloud base Range
CDF
Titre de la présentation - CONFIDENTIAL 26/05/2014 - 4 03/11/2015 - 4 TOPROF WC5 Meeting, Toulouse
10 20 30 40 50 60 70 80 90
720
730
740
750
760
770
780
790
Freq MHZ
SignalFit MLEBruit
Level 0
Level 1 Signal to Noise Ratio
Radial windspeed Dispersion
Horizontal wind speed Wind direction
Turbulence Windshear
Aerosol/cloud related data
Level 2
Temporal data
Data processing chain
Spectra
Titre de la présentation - CONFIDENTIAL 26/05/2014 - 5 03/11/2015 - 5 TOPROF WC5 Meeting, Toulouse
Numerous validations at reference independent sites (DTU, DEWI, Deutsche WindGuard…) against cup anemometers
Accuracy is 0.1 m/s on windspeed and 2° on wind direction
LIDAR profiler in DBS Mode: accuracy
Titre de la présentation - CONFIDENTIAL 26/05/2014 - 6 03/11/2015 - 6 TOPROF WC5 Meeting, Toulouse
Aerosol/cloud layer detection and discrimination
Attenuated backscatter coefficient retrieval
New aerosol/cloud features
Planetary Boundary Layer height detection
Mass concentration estimation (still in progress)
Titre de la présentation - CONFIDENTIAL 26/05/2014 - 7 03/11/2015 - 7 TOPROF, MC5, Toulouse
Aerosol/cloud layer detection
Titre de la présentation - CONFIDENTIAL 26/05/2014 - 8 03/11/2015 - 8 TOPROF WC5 Meeting, Toulouse
Detection of layers using CNR gradients
Step 1 : filtering and gradient calculation
Step 2 : Automatic gradient segmentation (iterative OTSU method 1979)
Step 3 : Base and top detection
CNR
Gradient
Binary mask
Base/top layers
Aerosol/cloud layers detection Principle
Titre de la présentation - CONFIDENTIAL 26/05/2014 - 9 03/11/2015 - 9 TOPROF WC5 Meeting, Toulouse
Relative error compared with reference structure depth (%)
Criti
cal S
ucce
ss In
dex
Refe
renc
e Al
gorit
hm re
sults
73% with less than 10% relative error
Aerosol/cloud layers detection Validation
Comparison with a reference database
Titre de la présentation - CONFIDENTIAL 26/05/2014 - 10 03/11/2015 - 10 TOPROF WC5 Meeting, Toulouse
Detection capabilities
CNR without filtering
Rang
e (m
) Ra
nge
(m)
Rang
e (m
)
CNR threshold = -27dB
CNR threshold = -28dB CNR threshold = -29dB
CNR threshold = -30dB CNR threshold = -31dB
Need to adjudt SNR thresholds defined for wind measurements for clouds/aerosols detection
Scenario ds dt
LOS 50m 1s
Titre de la présentation - CONFIDENTIAL 26/05/2014 - 11 03/11/2015 - 11 TOPROF, MC5, Toulouse
PBL detection
Residual layer
Free troposhere
Convective layer
Convective layer
Stable layer
Sunset Sunrise
Titre de la présentation - CONFIDENTIAL 26/05/2014 - 12 03/11/2015 - 12 TOPROF WC5 Meeting, Toulouse
Planetary boundary layer height Principle of the detection
Signal reflectivity
Gradient Gradient mask
Residual layer detection
Residual layer
Mixing layer
Threshold 0.2m/s [Bianco et al, 2013]
Vertical wind speed
Variance of the vertical wind speed
Mixing layer height
Binarisation of the variance image
Titre de la présentation - CONFIDENTIAL 26/05/2014 - 13 03/11/2015 - 13 TOPROF WC5 Meeting, Toulouse
Comparison with database from various campaigns
PBL height detection with a relative error <20% :
94% for the residual layer 78% for the mixing layer
Refe
renc
e Al
gorit
hm re
sults
PBL height detection Validation
Residual layer Convective layer Vertical pointing measurements (∆z = 25m, ∆t=2min)
Titre de la présentation - CONFIDENTIAL 26/05/2014 - 14 03/11/2015 - 14 TOPROF, MC5, Toulouse
Attenuated backscatter coefficent
Titre de la présentation - CONFIDENTIAL 26/05/2014 - 15 03/11/2015 - 15 TOPROF WC5 Meeting, Toulouse
Carrier-to-Noise Ratio (CNR)
Relative attenuated backscatter coefficient
Instrumental function calibration
Attenuated backscatter coefficient Principle
Automatic calibration is performed with low elevation PPI
A
y0
z0
w
Lorentzian shape with 4 parameters :
Titre de la présentation - CONFIDENTIAL 26/05/2014 - 16 03/11/2015 - 16 TOPROF WC5 Meeting, Toulouse
Comparison of WINDCUBE200S attenuated backscatter coefficient @1540nm with R-MAN510 absolute backscatter coefficient @ 355nm
Relative attenuated backscatter @ 1540 nm WINDCUBE200S [log10(m-1.sr-1)]
Absolute backscatter @ 355 nm R-Man510 [log10(m-1.sr-1)]
Hour Scenario ds dt
LOS 25m 1s
Attenuated backscatter coefficient Validation
Titre de la présentation - CONFIDENTIAL 26/05/2014 - 17 03/11/2015 - 17 TOPROF, MC5, Toulouse
Building networks of Doppler LIDARs
Titre de la présentation - CONFIDENTIAL 26/05/2014 - 18 03/11/2015 - 18 TOPROF WC5 Meeting, Toulouse
Why building networks of LIDARs ?
Enhance observation networks with Doppler LIDARs 1. To improve the « weather » monitoring
Wind hazards: Severe weather, Wind Shears, Aerosol hazards (plumes dispersion): Air pollution, Mining, industrial risks
Titre de la présentation - CONFIDENTIAL 26/05/2014 - 19 03/11/2015 - 19 TOPROF WC5 Meeting, Toulouse
Why building networks of LIDARs ?
Enhance observation networks with Doppler LIDARs 2. To improve weather nowcasts / forecasts
Area of interest? Surface / Ekman Layers (up to 200m / 500m) Boundary Layer (up to 1 – 2km) Troposphere (up to XX km)?
Spatial and temporal Scales of the met phenomena 10’, 2’?
Quantities to measure? Averaged Wind Speed / Direction Turbulence (tke?) Detection of Cloud / Aerosol layers ? Backscatter profiles?
Data of interest for the models? Define accuracy, resolution of required observations
Titre de la présentation - CONFIDENTIAL 26/05/2014 - 20 03/11/2015 - 20 TOPROF WC5 Meeting, Toulouse
Technical constraints for the LIDARs in networks
Reliable (MTBF, MTTR) Design to maintenance / to cost Tracability of the system during its entire life
Calibration / verification process Consistency of the data for each system
Determine accurately noise level Calibrate frequency offset Calibrate the correspondance between temporal signal and range gates Verify the accuracy of wind speed against a reference LIDAR
Titre de la présentation - CONFIDENTIAL 26/05/2014 - 21 Titre de la présentation - CONFIDENTIAL 18/09/13 - 21
UFO Project: Assimilation of LIDAR data into HARMONIE Model Set-up of Harmonie model domain for Toulouse
1km resolution: 500 x 500 grid points Calculation of Model background error Characteristics observations assimilation (area 400x400 grid points) Preprocessing set up for UFO observations
Radar/Lidar Scanners and Profilers TUBS aircraft data and Mode-S EHS data
Mode S EHS & MRAR versus HIRLAM Nestor Nested
Grid 200x200 5km 500x500 1km
Cycle 3 hours 1 hour
Lateral boundaries ECMWF hourly Nestor 15 min
Observations Temp/aircraft/ synop
Temp/aircraft/ synop
Assimilation 3DVAR 3 hour window
3DVAR 1 hour window
Courtesy of S. De Haan / KNMI
Titre de la présentation - CONFIDENTIAL 26/05/2014 - 22 03/11/2015 - 22 TOPROF WC5 Meeting, Toulouse
UFO Project: Assimilation of LIDAR data into HARMONIE Model
Experiments Conventional obs MF: coMFhr + Mode-S MRAR : mrar + LeoSphere WindcubeV2: vlmr + LeoSphere 3D wind observations
Requires additional QC
Comparison with WindCubeV2
Experiment vlmr: best performance Positive impact on wind speed standard
deviation in the first hour Wind speed bias is reduced in the first 2
hours
Courtesy of S. De Haan / KNMI
Titre de la présentation - CONFIDENTIAL 26/05/2014 - 23 03/11/2015 - 23 TOPROF WC5 Meeting, Toulouse
PBL wind, aerosol and clouds profiles assimilation into NWP models
Monitoring and nowcasting weather at a regional scale
http://www.nysmesonet.org/
125 standard sites 15 sites equipped with LIDAR Profilers & RADIOMETERs
> 100 000 km2 - PBL profilers < 5 000 km2 - covering typically a Megapole area - Long range scanning LIDARs
Weather monitoring and forecasting LIDAR networks
~1700 LIDARs needed !
Titre de la présentation - CONFIDENTIAL 26/05/2014 - 24 03/11/2015 - 24 TOPROF WC5 Meeting, Toulouse
Conclusion / Perspectives
Coherent doppler lidars are used operationnally in many applications : Wind energy, aviation weather, air quality, weather & climate…
Proven accuracy of the radial and reconstructed wind measurements (like in DBS mode)
Aerosol / cloud features can be retrieved from Dopler LIDARs Should be evaluated within TOPROF research working group: cloud base, backscatter; PBL
Long term assimilation of Doppler LIDAR data should be performed in TOPROF to evaluate the improvements of local weather forecasts
Titre de la présentation - CONFIDENTIAL 26/05/2014 - 25 03/11/2015 - 25 TOPROF WC5 Meeting, Toulouse
Current LIDAR products
Titre de la présentation - CONFIDENTIAL 26/05/2014 - 26 03/11/2015 - 26 TOPROF, MC5, Toulouse
Questions ?
Titre de la présentation - CONFIDENTIAL 26/05/2014 - 27 03/11/2015 - 27 TOPROF, MC5, Toulouse
Backup slides
Titre de la présentation - CONFIDENTIAL 26/05/2014 - 28 03/11/2015 - 28 TOPROF WC5 Meeting, Toulouse
A suite of atmospheric parameters
Wind
Turbulence Turbulence intensity (TI) Turbulent Kinetic Energy (TKE)
Radial Wind Speed Vertical wind speed Horizontal wind speed Wind direction
Aerosol/ Clouds
Structures Backscatter profile
Titre de la présentation - CONFIDENTIAL 26/05/2014 - 29
Titre de la présentation - CONFIDENTIAL 26/05/2014 - 30 03/11/2015 - 30 TOPROF WC5 Meeting, Toulouse
Choice of the wavelength of 1.5µm (NIR)
30
0.2
0.4
0.6
0.8
1
λ (µm) 1 10 2 3 4 5 6 7 8 9 15 0.8 0.6
I I II II III
Atmospheric Transmission
1e-07
1e-05
0.001
0.1
10
1 5 10 0.5
Exposition de 100 ns
Eye Safety
Er,Yb Yag
Tm-Ho Yag
Nd:Yag CO2
Diodes Laser
1.5µm
At 1.5µm:
- Availability of reliable laser sources
- High power Lidars are eye safety
- Transmission of lidar signal in atmosphere is very efficiency
Titre de la présentation - CONFIDENTIAL 26/05/2014 - 31 03/11/2015 - 31 TOPROF WC5 Meeting, Toulouse
Benefits of fiber technology
Why are we using fibers ? For guiding the light: avoid disalignment For amplifying the light: low consumption and reliable laser diodes
Laser
ν
Detector
AMPLIFIER
Telescope(s)
• Used of reliable optronic components of mainstream telecommunication market Low cost, reliable and compact architecture
• Flexible architecture that allows to adjust • Pulse Repetition Frequency (ie. Power) Measurement Range • Pulse length Spatial resolution
Titre de la présentation - CONFIDENTIAL 26/05/2014 - 32 03/11/2015 - 32 TOPROF WC5 Meeting, Toulouse
Long Range Scanning LIDARs
A 6 months trial has been performed in Palaiseau, France in 2014
Determine ranges for clear air conditions (visibility >10km, no rain) following the recommendations of ISO working groups on Doppler LIDARs Assess the impact of weather conditions on range
Measurements performed with a Scanning LIDAR WINDCUBE400S
Availability 80%
Availability 50%
Range 9km 11.5km
Titre de la présentation - CONFIDENTIAL 26/05/2014 - 33 03/11/2015 - 33 TOPROF WC5 Meeting, Toulouse
Long Range Scanning LIDARs
LIDARs measure at nominal performances in clear air conditions, ie. above 10km of visibility and with no rain For an all weather observations system, LIDAR must be coupled with RADAR Ra
nge
Rang
e
Rang
e
Visibility Cloud base
Rain rate
LIDARs measure at nominal performances in clear air conditions, ie. above 10km of visibility and with no rain For an all weather observations system, LIDAR must be coupled with RADAR
Titre de la présentation - CONFIDENTIAL 26/05/2014 - 34 03/11/2015 - 34 TOPROF WC5 Meeting, Toulouse
Short range LIDAR profiler
Operations in DBS Mode One sequence performed every 4s Worldwide 500 units are deployed
Measurements Performances
Range 40 to 290m
Number of heights 12
Data sampling rate 1s
Speed range 0 to 60 m/s 3 wind components (u,v,w)
Titre de la présentation - CONFIDENTIAL 26/05/2014 - 35 03/11/2015 - 35 TOPROF WC5 Meeting, Toulouse
Range resolution
Distance
Time
t1 0
Titre de la présentation - CONFIDENTIAL 26/05/2014 - 36 03/11/2015 - 36 TOPROF WC5 Meeting, Toulouse
Range resolution
Distance
Time
0
If pulse is very short, Spatial resolution is
t1 t1+tm Range gate
Titre de la présentation - CONFIDENTIAL 26/05/2014 - 37 03/11/2015 - 37 TOPROF WC5 Meeting, Toulouse
Range resolution
Distance
Time
0 t1 t1+tm Range gate
If pulse is square, range Resolution becomes:
Titre de la présentation - CONFIDENTIAL 26/05/2014 - 38 03/11/2015 - 38 TOPROF WC5 Meeting, Toulouse
Long Range Scanning LIDARs
Deployment during 3 months of a WINDCUBE200S at the certification test site in Denmark Objectives: To evaluate accuracy (mean difference) and precision (RMSE) of radial wind speeds and horizontal wind speeds
-15 -10 -5 0 5 10-15
-10
-5
0
5
10
Tower Radial Velocity (ms-1)
Lid
ar R
adia
l Vel
ocity
(ms-1
)
y = 0.99*x - 0.18R2 = 99.7%
RadialLinear Fit
116m met mast
Titre de la présentation - CONFIDENTIAL 26/05/2014 - 39 03/11/2015 - 39 TOPROF WC5 Meeting, Toulouse
Why building networks of LIDARs ?
Source ON THE’ KINETIC ENERGY SPECTRUM NEAR THE GROUND, A. OORT and A. TAYLOR, MONTHLY WEATHER REVIEW, VOLUME 97, NUMBER 9, SEPTEMBER 1969