Transcript Recommended specification of a Cloud Observing Station

Definition of instruments
for a cloud observing
station.
Specification for cloud
observing station.
• Fundamental instruments.
– Vertically pointing Dopplerised mm wave radar.
– Vertically pointing lidar.
– Multiple frequency vertically pointing microwave radiometer.
• Operating conditions.
– All weather. 24h/365days
– Resume reliable operation after rainfall.
– Unmanned – but visits during the working week.
• Data from operational forecast model.
– Height of freezing level, T and P profile for categorisation.
– Humidity & temperature profile - correct 94GHz attenuation.
– Horizontal wind (for turbulence) – or local wind profiler?
Specification for cloud
observing station.
• Fundamental instrument limits.
– Low level water clouds extinguish the lidar – upper clouds not seen.
– Radar may not detect water clouds, or ice clouds with small Xls.
• Sampling during rainfall.
–
–
–
–
Radomes wet during rain – unreliable data - quantifiable attenuation.
Response of normal rain gauges not rapid enough.
Rainfall CAN be detected from the Doppler signal of the falling drops.
To account for the lack of data during precip, when comparing with
model clouds, the model must be sub-sampled to remove rain periods
– Use operational radars to get IWC from Z in precipitation.
Basic and
Advanced Specification .
• Basic – for CloudNET analysis – cloud fraction and
water content.
– mm wave Doppler radar.
– Ceilometer to detect water clouds.
– Dual frequency microwave radiometers for LWP.
• More advanced cloud station.
– High sensitivity mm wave Doppler radar
– Sensitive lidar.
– Microwave radiometer for LWP and humidity profiling.
Cloud radar : frequency?
• 35Ghz
– Mature technology
– Low attenuation by oxygen, water vapour and liquid water
clouds.
– Large insect return. Recognise using Doppler.
• 94GHz.
– New technology. Tube lifetime?
– Attenuation by O2 & water vapour  2dB: liquid 1dB /100g/m-2.
– Low return from insects.
• 10Ghz (X-band)
– Low cost – cheaper components.
– Sufficient sensitivity?
Performance of 35GHz tube.
Adjust
drive
voltage
1.5dB
Nov 04
Oct 05
Performance of three 94GHz tubes
>10dB loss
11 months
Conclusion – Use 35GHz
Pulse coding:
8bit pulse coding
10dB more sensitive
than single pulse.
No range side lobes.
Z errors up to 1dB if
target high Doppler and
high Doppler width.
i.e only in ppn.
Radar sensitivity?
Cabauw: large antenna,
-55dBZ at 1km for 30secs and 60m. Sufficent?
STRATOCUMULUS: always seen by lidar.
9200 hours – cumulative frequency.
at –50dBZ still see only 56% of Sc.
BUT can always
rely on simple
lidar to see Sc.
Radar sensitivity: ice clouds?
Use molecular backscatter from sensitive lidar to find optical depth,
want to see  =0.05: is cloud detected by radar?
Sensitivity Z=-45dBZ at 1km
miss some clouds with  =0.05
Sensitivity –55dBz at 1km
OK. Miss v few low  clouds
Radar sensitivity v cost.
• Sensitivity Z=-45dBZ @ 1km is useful
– misses some high clouds.
• Sensitivity -55dBZ at 1km ideal
This is achieved with commercial systems using
coded pulse. Cost is about 500K euro.
(magnetron or coded pulse??)
• Studies suggest FM/CW system would be cheaper.
• With careful design minimal range sidelobes.
• Sensitivity of –50dBZ at 1km for about 100,000 euro?
• Needs to be studied further.
• X-band possibility.
LIDAR SENSITIVITY: cloudnet algorithm
only needs lidar to see water clouds – very easy with simple system.
Cost 20,000 Euro
MICROWAVE RADIOMETER
•Current systems need Tb to 0.5K – 150,000 euro
•Calibration with black bodies, tip curves.
•Relax the specification - tolerate 5K drift in Tb?
•Must be reasonably stable during cloudy periods.
•What would be the price?
50,000 Euro??
BASIC SPEC:
radar (-45dBZ@1km)100K,
ceilometer
20K
radiometers
50K
- total 170K-200K??
+ P, T, q and wind profiles from model
or wind from local profiler.
Z in raining ice clouds from operational radar
At this price met services could install a network,
initially for model evaluation, ultimately for data
assimilation.
Lidar high sensitivity:
How sensitive a lidar to see thin high ice clouds? Several available:
532nm/ 10km molecular 6 10-7 /m/sr with snr 6: i.e.  10 –7/m/sr
Comparison of fractional
cloud cover inferred from
radar -45kdBZ at 1 km
and sensitive lidar
(10 –7/m/sr at 10km.
SOLID line observed
Dotted line – model
NEED GOOD LIDAR
TO SEE ICE IN THE MODEL
High Sensitivity Lidar.
• uv or green - sensitivity 10 –7/m/sr at 10km.
• Sees thin ice clouds with low optical depth.
• Use molecular return – for unambiguous optical depth.
• Also good for profiling aerosols
direct optical depth of aerosols
• depolarisation channel
for shape of ice crystals and aerosols
Cost of new systems now being advertised:
weather proof
Unmanned: 150,000 euro?
Need to demonstrate their performance in the field.
High Sensitivity Specification
•Radar –55dBZ @ 1km: 500k
•Lidar
: 150k
•Profiling Radiometers
150k
TOTAL
800k???
SUITABLE FOR A FEW SPECIALISED SITES.