Application of the dual-beamwidth method to a narrow beam MF radarfor estimation of spectral width

R. Latteck, , N. EnglerW. SingerLeibniz-Institut für Atmosphärenphysik, Schloss-Str. 6, D-18225 Kühlungsborn, Germany10th International Workshop on Technical and Scientific Aspects of MST Radar

Piura, Peru, May 13-20, 2003

IntroductionSpectral widths observed by narrow beam VHF/UHF Doppler radars are usedto estimate turbulent energy dissipation rates. In case of broader beams, theobserved spectral widths have to be corrected for the influence of beam andshear broadening using simultaneously measured horizontal winds. VanZandtet al. developed a new dual-beam width method to estimate the turbulent com-ponent of spectral width from MST radar observations without any additionalassumptions and tested it successfully for the troposphere.

In summer 2002 the new SauraMFradarwasputintooperationontheAndoyaisland in Norway.The system has high flexibility in antenna beam forming allo-wing off-zenith beams with different beam widths. In addition, the beam stee-ring capabilities of the Saura MF radar and the nearby located ALWIN VHFradar provide common volume observations at mesospheric altitudes in sum-merduring the appearance of PMSE.

Experiments with different beam widths have been carried out with the MFradar to test the dual-beam width method at mesospheric altitudes. We comparespectral width estimates from both the single-beam width and the dual-beamwidth method on a case study basis.

The method

The Saura MF radar

z

The (2BW) considers that the dominant terms indual-beamwithmethodσ θ σ

θ σθ θ

σ σ

2 2 2

2 2

2 2

beam+shear corr

corr

n w

obs,n obs,w

are proportional to what means that is also approximately pro-portional to . If is measured simultaneously in nested volumes with anarrow beamwidth and a broad beamwidth two simultaneous equationsfor and can be solved, what results in:

The variance of the observed radar Doppler spectral width is related tovelocity variance due to turbulence and influenced by andwhat are variances due to the interaction of the background wind and shear andof waves with the radar beam:

σ2obs

σ σ σ2 2 2turb beam+shear wave

The standard (1BW) consists of estimatingevaluating

single-beamwidth method σσ

2

2wave

beam+shearand by

and substracting σ σ σ2 2 2corr obs turbfrom to get .

Tx/Rx2kW

Tx/Rx2kW

Tx/Rx2kW

Tx/Rx2kW

+

+

DDS

DDS

DDS

DDS

Envelope

1

2

3

4

cntrl 1...16

Controler module 1

Tx/Rx2kW

Tx/Rx2kW

Tx/Rx2kW

Tx/Rx2kW

+

+

DDS

DDS

DDS

DDS

Envelope

61

62

ref

ref

ref

Controler module 16

1 2 31…

Transceiver(4 channel)

Controller

Synthesiser& Reference

Generator

IF CombinerRx4

Rx3

Rx2

Rx1

DA

DA

DA

DA

cntrl

Control Analysis

Transmitter5.5°

NMills CrossAntenna

IP1

IP2

IP3

IP4

QP1

QP2

QP3

QP4

Tx/Rx2kW

Tx/Rx2kW

Tx/Rx2kW

Tx/Rx2kW

+

+

DDS

DDS

DDS

DDS

Envelope

1

2

3

4

cntrl 1...16

Controler module 1

Tx/Rx2kW

Tx/Rx2kW

Tx/Rx2kW

Tx/Rx2kW

+

+

DDS

DDS

DDS

DDS

Envelope

61

62

ref

ref

ref

Controler module 16

1 2 31…

Transceiver(4 channel)

Controller

Synthesiser& Reference

Generator

IF CombinerRx4

Rx3

Rx2

Rx1

DA

DA

DA

DA

cntrl

Control Analysis

Transmitter5.5°

NMills CrossAntenna

IP1

IP2

IP3

IP4

QP1

QP2

QP3

QP4 SAURA MF RADARSIMPLIFIED BLOCK DIAGRAM

Radiation pattern plots of Saura MF radar antennafor a 17.2° tilded narrow (red) and broad (blue) beamto northwest direction. The red and blue lines in thecontour plots mark the plane of the slices through the3D radiation diagram shown obove.

-40

-35

-30

-25

-20

-15

-10

-5

0

-1 -0.5 0 0.5 1-1

-0.5

0

0.5

1

sin( ).cos( )θ ϕ

N

E

-40

-35

-30

-25

-20

-15

-10

-5

0

-1 -0.5 0 0.5 1-1

-0.5

0

0.5

1

sin( ).cos( )θ ϕ

N

E

Radiation pattern (top view)of tilded narrow beam

Radiation pattern (top view)of tilded broad beam

-30 -20 -10 0 10 20 300

2

4

6

8

10

12

14

16

18

20

22

θ[°]

Radiation pattern ( = 320.5 / =17.2°)ϕ θ

19.25 dB

15.82 dB∆θ-3dB = 6.6°

∆θ-3dB = 13.8°

-17.2

Advantage:neither the coefficients in equation (2) nor the windspeeds and shears appear in equations (3) and (4)

σθ σ θ σ

θ θturbw obs n n obs w

w n

22 2 2 2

2 2=⋅ − ⋅

−, ,

σ θσ σ

θ θcorr n nobs w obs n

w n,

, ,2 22 2

2 2=−−

σθθ

σcorr ww

ncorr n, ,

22

22=

⋅

(3)

(4)

(5)

σ σ σ σ σ σobs turb beam shear wave turb corr2 2 2 2 2 2= + + = ++ (1)

( ) ( ) ( )[ ]σθ

α α α αbeam shear z zU V U u R u R+ = + − +2 2 2 2 2 4 2

32cos cos sin sin (2)

Radar frequency: 3.17 MHzPeak power: 116 kWMean power:(0.2% dc):Pulse form: Gauss

Pulse width: > 7 sRange resolution: 1000 mAntenna: 29 crossed

half wave dipolsHalf power beam width:6,4°Beam directions: vertical,

off-zenith

230 W

µ

The experimentThe 29 crossed dipols of the new Saura MF radar areconnected to individual transceiver modules each. Ifless than 29modules are used a broad antenna beam isformed. A sequence of narrow beam ( ) andbroad beam experiments, each pointinginterleaved with towards NW andSE, were run in April 2003 to test the dual beam me-thod. Additionally 4 DBS narrow beam experimentsat 7.3° off-zenith provided the horizontal wind infor-mation.

( )17.2° off-zenith

θθ

HPFW

HPFW

=6.6°=13.8°

06:00 09:00 12:00 15:00 18:00

60

70

80

90

SE (140.5°) narrow beam

SE (140.5°) broad beam06:00 09:00 12:00 15:00 18:00

60

70

80

90

NW (320.5°) narrow beam

NW (320.5°) broad beam06:00 09:00 12:00 15:00 18:00

60

70

80

90

06:00 09:00 12:00 15:00 18:00

60

70

80

90

23-Apr-2003 [UT]

Radial velocity[m/s]

-10

0

10

[m/s]

-10

0

10

[m/s]

-10

0

10

[m/s]

-10

0

10

06:00 09:00 12:00 15:00 18:00

60

70

80

90

SE (140.5°) narrow beam

SE (140.5°) broad beam

[dB]

40

60

80

100

06:00 09:00 12:00 15:00 18:00

60

70

80

90

NW (320.5°) narrow beam

NW (320.5°) broad beam

[dB]

40

60

80

100

06:00 09:00 12:00 15:00 18:00

60

70

80

90

[dB]

40

60

80

100

06:00 09:00 12:00 15:00 18:00

60

70

80

90

23-Apr-2003 [UT]

[dB]

40

60

80

100

Signal power

06:00 09:00 12:00 15:00 18:00

60

70

80

90

SE (140.5°) narrow beam

SE (140.5°) broad beam06:00 09:00 12:00 15:00 18:00

60

70

80

90

NW (320.5°) narrow beam

NW (320.5°) broad beam06:00 09:00 12:00 15:00 18:00

60

70

80

90

06:00 09:00 12:00 15:00 18:00

60

70

80

90

23-Apr-2003 [UT]

Spectral width[m/s]

0

5

10

[m/s]

0

5

10

[m/s]

0

5

10

[m/s]

0

5

10

ReferencesVanZandt,T.E., G.D.Nastrom, J. Furumoto, T. Tsuda, and W.L. Clark, A dual-beamwidth radar method for measuring atmospheric turbulent kinetic energy,Geophys. Res. Lett., Vol. 29,No.12,13-1-13-3,2002Lübken, F.-J., The thermal structure of the Arctic summer mesosphere,

, 104, 9135-9149, 1999.Müllemann, A., M. Rapp, F.-J. Lübken, and Peter Hoffmann, In situmeasurements of mesospheric turbulence during spring transition of theArcticmesosphere,

J.Geophys. Res.

Geophys. Res. Lett., Vol. 29, No. 10, 115-1 - 115-4, 2002.

NW 14° NE 14°

SW 14°SE 14°

NW 7°

SW 7° SE 7°

NE 7°

W 17,2°

S 17,2°

E 17,2°

N 17,2°

W 7,3°

S 7,3°

N 7,3°

E 7,3°

0 10 20 30 40 50 50km

Future common MF/VHF experimentsThe high flexible beam steering of the Saura MF radar allowscommon volume observations at mesospheric altitudes with theSoutheast and Southwest beams of the

.All pos-sible beams of the ALWIN VHF radar and most of the possibleSaura MF radar beams point also into the volume that is illumi-nated by the Andenes MF radar as well. This makes it possibleto compare results obtained from the same volume but withradars operating on different frequencies and using similar ordifferent analysis techniques. Furthermore

f the Saura MF radar beamis steered to the volume where the ALWIN NW 7° or NW14°beam point to, the rocket trajectories have intersections withboth radar beams at altitudes between 75 and 90 km.

nearby locatedALWINVHF radar andwiththelidarexperiments atALOMAR

comparisons within-situ measurements from sounding rockets launched from theAndoya Rocketrange are possible. I

Areas byoblique and vertical beams ofALWINVHF radar (green),AndenesMFradar(blue) and

at 85km altitude illuminated

SauraMFradar(red).Northern part of Andoya island.

0 1 2 3 4 5km

ALWINVHF radar

AndenesMF radar

SauraMF radar

ALOMAR

AndoyaRocket Range

Experimentconfigurations broad narrow DBS

beam beam windPRF [Hz] 150 150 150Coherent integrations 16 16 8Data points per beam 512 512 800Range start [km] 50 50 50Range end [km] 98 98 98Sampling interval [km] 2 2 2Pulse width [ s] 26.67 26.67 26.67Beam direction sequence 140.5/17.2 140.5/17,2 50.5/7.3

(azimuth/zenith) 320.5/17.2 320.5/17.2 230.5/7.3140.5/7.3320.5/7.3

Time resolution t [ms] 213.33 213.33 213.33Experiment run time [s] 109.23 109.23 170.67

µ

∆

ResultsThe dual-beamswereformed at 17.2° ze-nith angle with a minimum of the antennaradiation pattern in vertical direction to re-duce the contamination by specular re-flections. Individual estimates of turbu-lent spectral width are obtained from con-secutive narrow and broad beam obser-vations every 8 minutes with interleavedNW/SE observations for the narrow andbroad beam. The 3-hour mean profile of

shows velocity variances between 5 and about 15 m /s below 78 km where contributions from specular reflections should be small or negligi-ble. These turbulent velocity variances correspond to energy dissipation rates between 10 and 60 mW/kg based on a mean temperature profile forApril after falling sphere measurements atAndenes (Lübken, 1999). The estimated dissipation rates are in agreement with rocket observations du-ring spring transition (Müllemann et al., 2002). During the first dual-beamwidth observations the wind speeds were small (less than 10m/s)andthecorrection for beam and shear broadening used in the single-beamwidthmethod is negligible. (Note: different abscissa scales in the middle panels!)

σ2 2 2turb

Contour plots of signal power, radial velocity, and spectral width obtained with narrow and broad beam experiments at 17.2° solar zenith angle.The differences in signal power between narrow and broad beams are due to less numbers of antennas forming the broad beam. The effectiveradiated power is reduced by about 50 per cent. The radial velocities in the middle panel show reversed signs for the opposite directed narrow andbroad beams. For the broad beams lager spectral widths are observed all the time (right panel).

σ2obs

2 2(m s- ) σ2corr

2 2(m s- ) σ2turb (m s- )2 2

0 10 20 3064

68

72

76

80

84

88

SENW

0 2 4 6 8 10 0 10 20 3064

68

72

76

80

84

88

NW SE

23-Apr- 200310:47 - 13:56 UT

Dual-Beamwidth Spectral-Width Method (VanZandt et al., 2002)( narrow beam, HPHW=3.4°, SZA=17° / ------ broad beam, HPHW=7°, SZA=17°)

Single-Beamwidth Spectral-Width Method( narrow beam, HPHW=3.4°, SZA=17°)

0 10 20 3064

68

72

76

80

84

88

SENW

0.0 0.5 1.0 0 10 20 304

68

72

76

80

84

8823-Apr-200310:47 - 13:56 UT

NW SE

σ2obs

2 2(m s- ) σ2corr

2 2(m s- ) σ2turb (m s- )2 2