Measuring Soil Water Content with Ground Penetrating Radar: A Review
Measuring Snowfall with a Low-Power K-Band Radar (Micro ...Measuring Snowfall with a Low-Power...
Transcript of Measuring Snowfall with a Low-Power K-Band Radar (Micro ...Measuring Snowfall with a Low-Power...
Measuring Snowfall with a Low-Power K- Band Radar (Micro Rain Radar) in Polar Regions
University of Cologne
Institute for Geophysics and Meteorology
Maximilian Maahn, [email protected]
Maximilian MaahnP. Kollias, S. Kneifel, I. Gorodetskaya, G. Peters, C. Simmer
27/06/12
ERAD 2012, Toulouse
1
Micro Rain Radar (MRR)
• Vertically pointing• 24 GHz (12.4 mm)• 30 range gates• 30...200 m vertical
University of Cologne
Institute for Geophysics and Meteorology
Maximilian Maahn, [email protected] University of Cologne
• 30...200 m vertical resolution
• FM-CW• 50 mW, 25 W, 20 kg
• Developed for rain2
27/06/12
ERAD 2012, Toulouse
Can we use MRRs for snowfall?
University of Cologne
Institute for Geophysics and Meteorology
Maximilian Maahn, [email protected] University of Cologne
3
27/06/12
ERAD 2012, Toulouse
Get Ze from MRR data
Good performance > 3 dBz
Must not be used for snow, because Z is derived from DSD assuming only rain
Integrated Spectra (Z, W…)
University of Cologne
Institute for Geophysics and Meteorology
Maximilian Maahn, [email protected]
27/06/12
4
ERAD 2012, Toulouse
Good performance > 3 dBz Incomplete noise removal disturbs smaller reflectivities. Correction of aliasing effects needed (Kneifel et al.,2011)
Gives best results for snow with proposed routine→ Sensitivity ~ -5 dBz
Raw Doppler Spectra
Noise & Height corrected Doppler Spectra
New routine based on Raw Doppler Spectra
1. Determine the most significant peak including its borders (modified Hildebrand and Sekhon, 1974)
2. Define remaining spectrum as noise and remove noise from peak
19/0
1/20
12, 2
3:12
Standard scheme New scheme
Hei
ght [
m]
University of Cologne
Institute for Geophysics and Meteorology
Maximilian Maahn, [email protected]
5
27/06/12
ERAD 2012, Toulouse
remove noise from peak3. Dealiase the spectrum,
because Nyquist range is small (using Ze - v relation by Atlas et al., 1973)
4. Calculate Z e, SNR, W
� Independent on type of hydrometeor!
Doppler velocity [m/s]
Hei
ght [
m]
New MRR routine is available as open source at http://gop.meteo.uni-koeln.de/software/
Comparison with Cloud Radar
• Reference: MIRA35 35.2 GHz cloud radar
• Collocated measurements January-April 2012
University of Cologne
Institute for Geophysics and Meteorology
Maximilian Maahn, [email protected]
April 2012 • at UFS
Schneefernerhaus(German Alps, 2650m)
• ∆t = 60s
6
27/06/12
ERAD 2012, Toulouse
MIRA vs. MRR: Z eStandard Product Proposed Routine
Offset due to different Frequencies
MR
R Z
e [d
Bz]
MR
R Z
[dB
z]
high spread
University of Cologne
Institute for Geophysics and Meteorology
Maximilian Maahn, [email protected]
7
27/06/12
ERAD 2012, Toulouse
MIRA 35 Ze [dBz]MIRA 35 Ze [dBz]Incomplete noise removal
MR
R Z
e [d
Bz]
MR
R Z
[dB
z]
MIRA vs. MRR: Frequency by altitude
MIRA35 MRR (new routine)
Hei
ght [
m]
Hei
ght [
m]
University of Cologne
Institute for Geophysics and Meteorology
Maximilian Maahn, [email protected]
Ze [dBz]
-5 dBz � 0.01 - 0.1 mm/h
Ze [dBz]
8
27/06/12
ERAD 2012, Toulouse
Hei
ght [
m]
Hei
ght [
m]
MIRA vs MRR: WStandard Product Proposed Routine
Noise
MR
R W
[m/s
]
MR
R W
[m/s
]
University of Cologne
Institute for Geophysics and Meteorology
Maximilian Maahn, [email protected]
9
27/06/12
ERAD 2012, Toulouse
MIRA35 W [m/s] MIRA35 W [m/s]
MR
R W
[m/s
]
MR
R W
[m/s
]
Outlook: Application to MRR measurements
• Ny-Ålesund (79 °°°°N): winter 2010/11
• Longyearbyen, (78 °°°°N): winter 2010/11
• Princess Elisabeth station (72 °°°°S): summer seasons 2010,11 and 12
University of Cologne
Institute for Geophysics and Meteorology
Maximilian Maahn, [email protected]
2010,11 and 12
27/06/12
10
ERAD 2012, Toulouse
(NASA)
Frequency by altitude diagrams
Ze [dBz]
Hei
ght [
m]
Spitsbergen
Antarctica
University of Cologne
Institute for Geophysics and Meteorology
Maximilian Maahn, [email protected]
27/06/12
11
ERAD 2012, Toulouse
Ze [dBz] Ze [dBz]
Hei
ght [
m]
Hei
ght [
m]
Conclusion
1) MRR standard products highly biased for snow.
2) But, snow can be observed (> -5 dBz) if the new scheme is used.
University of Cologne
Institute for Geophysics and Meteorology
Maximilian Maahn, [email protected]
if the new scheme is used.3) MRR has potential for snowfall
climatologies
12
27/06/12
ERAD 2012, Toulouse
Thank you for your attention
University of Cologne
Institute for Geophysics and Meteorology
Maximilian Maahn, [email protected]
Thanks to ERAD for covering travel expenses!
Maahn and Kollias (2012): Improved Micro Rain Radar Snow Measurements Using Doppler Spectra Post-Processing, submitted to AMT
New MRR routine is available at http://gop.meteo.un i-koeln.de/software/
27/06/12
13
ERAD 2012, Toulouse
University of Cologne
Institute for Geophysics and Meteorology
Maximilian Maahn, [email protected]
27/06/12
14
ERAD 2012, Toulouse
Back-up slides
University of Cologne
Institute for Geophysics and Meteorology
Maximilian Maahn, [email protected]
27/06/12
15
ERAD 2012, Toulouse
Comparison of sigma
University of Cologne
Institute for Geophysics and Meteorology
Maximilian Maahn, [email protected]
27/06/12
16
ERAD 2012, Toulouse
Comparison with Cloud Radar
• Reference: MIRA35 35.2 GHz cloud radar
• January-April 2012 at UFS Schneefernerhaus
University of Cologne
Institute for Geophysics and Meteorology
Maximilian Maahn, [email protected]
Schneefernerhaus (German Alps)
• >15% of data had to be discarded because of attenuation due to wet snow on dish
17
27/06/12
ERAD 2012, Toulouse
Metek's (indirect) approach for Z
Recorded Doppler Power Spectrum Z ~ D6, R ~ D3
P v( ) → P D( ) →N D( ) → Z, R
Assume Doppler velocity = fall velocity
Convert to DSD using Mie Theory
University of Cologne
Institute for Geophysics and Meteorology
Maximilian Maahn, [email protected]
• Works good for rain! Easy to compare with Z from C-band radar, Parsivel etc.
18
27/06/12
ERAD 2012, Toulouse
Assume Doppler velocity = fall velocityand convert to drop size usingAtlas et al. (1973)
Why does it not work for snow?Size – fall velocity different for snow and has much higher uncertainty
Z ~ D6, R ~ D3
not feasible for snow
P v( ) → P D( ) →N D( ) → Z, R
University of Cologne
Institute for Geophysics and Meteorology
Maximilian Maahn, [email protected]
�Z, R are biased for snow!�Instead: calculate Z e from Doppler
spectrum P(v)
Backscattering cross section differentfor snow and depends heavily on particle type
19
27/06/12
ERAD 2012, Toulouse
Doppler Spectrum cannot be used19
/01/
2012
, 23:
12
Noise is not properly removed from the spectra
University of Cologne
Institute for Geophysics and Meteorology
Maximilian Maahn, [email protected]
Sensitivity is lost due to contamination with noise
Aliasing due to limited unambiguous Doppler range
Z [dBz]
20
27/06/12
ERAD 2012, Toulouse
Flow chart diagram
University of Cologne
Institute for Geophysics and Meteorology
Maximilian Maahn, [email protected]
27/06/12
21
ERAD 2012, Toulouse
Get Zere
flect
ivity
[dB
]
University of Cologne
Institute for Geophysics and Meteorology
Maximilian Maahn, [email protected]
05/07/12
22
IRTG first year evaluation
V [m/s]
Spe
ctra
lref
lect
ivity
Get Zere
flect
ivity
[dB
]
University of Cologne
Institute for Geophysics and Meteorology
Maximilian Maahn, [email protected]
05/07/12
23
IRTG first year evaluation
V [m/s]
Spe
ctra
lref
lect
ivity
Ze
Get Zere
flect
ivity
[dB
]
University of Cologne
Institute for Geophysics and Meteorology
Maximilian Maahn, [email protected]
05/07/12
24
IRTG first year evaluation
V [m/s]
Spe
ctra
lref
lect
ivity
Ze
noise noise
Remove clutter
University of Cologne
Institute for Geophysics and Meteorology
Maximilian Maahn, [email protected]
05/07/12
25
IRTG first year evaluation
Resulting Spectra
University of Cologne
Institute for Geophysics and Meteorology
Maximilian Maahn, [email protected]
05/07/12
26
IRTG first year evaluation
Aliasing
Aliasing
University of Cologne
Institute for Geophysics and Meteorology
Maximilian Maahn, [email protected]
27/06/12
27
ERAD 2012, Toulouse
Virga
University of Cologne
Institute for Geophysics and Meteorology
Maximilian Maahn, [email protected]
27/06/12
28
ERAD 2012, Toulouse