1

Frequent rain observation from geostationary satellitesby millimetre-submillimetre-wave sounding

 Bizzarro Bizzarri, CNR Istituto di Scienze dell’Atmosfera e del Clima, Italy

Albin Gasiewski, NOAA Environmental Technology Laboratory, USADavid Staelin, MIT Research Laboratory of Electronics, USA

Content

• Requirements for frequent precipitation observation

• Principle of MW precipitation observation from geostationary orbit

• GOMAS (Geostationary Observatory for MW Atmospheric Sounding)

 WMO/CGMS 1st Internationl Precipitation Working Group Workshop, Madrid, 23-27 September 2002

2

Requi- G l o b a l N W P RegIonal N W P N o w c a s t i n g

rement Unit W M O E U M W M O E U M W M O E U M Target Threshold Target Threshold Target Threshold Target Threshold Target Threshold Target Threshold

x km 50 100 5 100 10 50 3 50 5 50 1 5

r.m.s. mm/h 0.1 1 0.1 1 0.1 1 0.1 1 0.1 1 1 10

t h 1 12 1 12 0.5 6 0.5 3 0.08 1 0.03 0.3

delay h 1 4 1 4 0.5 2 0.5 1 0.08 0.5 0.03 0.1

WMO and EUMETSAT requirements for precipitation observation

3

The problem of measuring precipitation 

The most long-standing practise makes use of VIS/IR image sequences from GEO:

• The observing frequency is suitable (15 min with MSG)

• The measurement is strongly indirect (VIS/IR only “sees” the cloud top)

• The applicability is mostly addressing convective precipitation. 

From LEO, MW images are used (from SSM/I, TRMM, in the near future GPM):• The measurement is direct for low frequencies (10 GHz), less for high frequencies (90 GHz)

• The applicability is better for convective precipitation, but extends to all precipitation types

• The frequency, with GPM, will be around 3 hours

• Global scale will be served optimally, regional scale to a fair extent

• For the mesoscale, fusion between LEO MW images and GEO VIS/IR images will be attempted. 

The ideal would be to extend the MW imagery technique to the geostationary orbit, but:• The antenna diameter for a 10-km resolution is 15 m at 90 GHz, 35 m at 37 GHz, 70 m at 19 GHz

• Polarisation diversity is not practical from the geostationary orbit (large and variable -angle). 

GEO requires using higher frequencies and exploiting a different physical principle.

4

The physical principle for measuring precipitation from GEO 

MW observation from LEO makes use of “atmospheric windows”:• Most common frequencies: 6 GHz, 10 GHz, 19 GHz, 23 GHz, 37 GHz, 90 GHz, 150 GHz• Dual polarisation for roughness effects (over the sea) and scattering from ice (over land). 

The proposed principle for GEO exploits absorption bands:• Profiles of temperature and humidity are measured by more bands at different frequencies• Profiles observed exploiting bands of different frequencies are differently sensitive to clouds• Absorption bands at very high frequencies enable using antennas of affordable size• It is not necessary to differentiate polarisations• In absorption bands the measurement is equally effective over sea and land.

The precipitation measurement passes through the sounding one, therefore:• Many more channels are needed, very narrow (< 1 ‰), with SNR > 100• In exchange, one simultaneously gets:

- the temperature vertical profile (also inside clouds)

- the humidity vertical profile (also inside clouds)

- the columnar content (or gross profile) of liquid water in the cloud

- the columnar content (or gross profile) of ice water in the cloud

- the precipitation.

5

Multi-channel image from an airborne radiometer (Gasiewski et al, 1994). Note:

• In absorption bands (183 and 325 GHz) increasing cloud impact moving from the peak to the window (from 183 1 to 183 3 and 183 7; and from 325 1 to 325 3 and 325 9);

• In windows (89, 150, 183 7, 220 e 325 9) increasing cloud impact with increasing frequency.

6

Precipitation images from a cold front on October 7, 1998: NEXRAD precipitation map smoothed to 15 km resolution (left image), and NOAA/AMSU precipitation map obtained using a neural net retrieval technique (right image) (Staelin and Chen, 2000).

7

NOAA-15 AMSU-B 183+/-7 GHz 15-km resolution imagery of small rain cells over the Eastern U.S on August 2, 2000 (from Staelin and Chen, 2000).

8

Comparison between the 118/54 GHz profile ratio from the NAST-M microwave radiometer being flown on the NASA ER-2 aircraft and simultaneous EDOP Doppler radar reflectivity observation. Hurricane Bonnie at 17 GMT on August 26, 1998 (Tsou et al, 2001). The ratio 118/54 is unit in the absence of precipitation or above the precipitation cell, and reduces in the presence of precipitation because of higher attenuation from drops and impact from ice scattering at the higher frequency. The agreement between the two vertical cross-sections is striking. Horizontal maps of this information from GEO at 15 min intervals, will resemble a proxy rain radar operating over continental field of view, particularly over oceans and mountainous terrain.

9

Atmospheric spectrum in the MW/Sub-mm range (Klein and Gasiewski, 2000).

Preferred bands - for O2 (temperature): 54 GHz, 118 GHz, 425 GHz; for H2O: 183 GHz, 380 GHz.

10

Incremental Weighting Functions (IWF) for the selected channels in the bands of oxygen (54, 118 and 425 GHz) and of water vapour (183 and 380 GHz). Note that the sensitivity of 380 and 425 GHz drastically drops in lower troposphere (Klein and Gasiewski, 2000).

11

Geophysicalparameter

Horizontal resolution

Vertical resolution

Accuracy (r.m.s.)

Observing cycle

Delay of availability

Temperature profile 30 km 2.0 km 1.5 K 1 h 15 min

Humidity profile 20 km 2.0 km 20 % 1 h 15 min

Cloud liquid/ice water 20 km 5.0 km 50 % 15 min 5 min

Precipitation rate 10 km - 5 mm/h 15 min 5 min

Antenna Ø 54 GHz 118 GHz 183 GHz 380 GHz 425 GHz

1 m 242 km 112 km 73 km 35 km 31 km

2 m 121 km 56 km 36 km 18 km 16 km

3 m 81 km 37 km 24 km 12 km 10 km

4 m 60 km 28 km 18 km 8.8 km 7.8 km

Reference user requirements adopted for GOMAS

Resolution v/s frequency and antenna diameter

12

GOMAS (Geostationary Observatory for MW Atmospheric Sounding)European sector (1/12 of disk) scanned each 15 minutesTemperature vertical profile, all-weather and inside clouds: resolution 30 kmHumidity vertical profile, all-weather and inside clouds: resolution 20 kmColumnar content or gross profile of liquid and ice water: resolution 20 kmPrecipitation: resolution 10 km.

13

(GHz)

(MHz)

Product resolution

IFOV

s.s.p.

Required NET (K)

Expected

NET (K)

Peak of IWF

(GHz)

(MHz)

Product resolution

IFOV

(s.s.p.)

Required NET (K)

Expected

NE T (K)

Peak of IWF

56.325 50     0.6 0.15 27 km 183.310  0.300 300     0.6 0.45 10 km

56.215 50     0.5 0.15 23 km 183.310  0.900 500     0.6 0.35 8.5 km

56.025 250     0.5 0.07 17 km 183.310  1.650 700 2 x 2 pixels   0.5 0.29 7 km

55.520 180     0.4 0.08 13 km 183.310  3.000 1000 (20 km) 24 km 0.3 0.24 6 km

54.950 300 6 x 6 pixels   0.4 0.06 10 km 183.310  5.000 2000     0.4 0.17 5 km

54.400 220  (60 km) 81 km 0.3 0.07 8 km 183.310  7.000 2000     0.6 0.17 4 km

53.845 190     0.3 0.08 5 km 183.310  17.000 4000     0.3 0.18 surface

53.290 360     0.3 0.06 3 km 380.197  0.045 30     0.3 2.36 15 km

52.825 300     0.2 0.06 2 km 380.197  0.400 200     0.5 0.91 13 km

51.760 400     0.1 0.05 1 km 380.197  1.500 500 2 x 2 pixels   0.5 0.58 11 km

50.300 180     0.1 0.08 surface 380.197  4.000 900 (20 km) 12 km 0.5 0.43 9 km

118.750  0.018 6     0.5 1.32 34 km 380.197  9.000 2000     0.4 0.29 7 km

118.750  0.035 12     0.6 0.93 29 km 380.197  18.000 2000     0.3 0.36 6 km

118.750  0.080 20     0.6 0.72 24 km 424.763  0.030 10     0.5 3.40 34 km

118.750  0.200 100     0.5 0.32 19 km 424.763  0.070 20     0.6 2.41 28 km

118.750  0.400 200 3 x 3 pixels   0.5 0.23 15 km 424.763  0.150 60     0.6 1.39 23 km

118.750  0.700 400 (30 km) 37 km 0.5 0.16 12 km 424.763  0.300 100 3 x 3 pixels   0.5 1.08 18 km

118.750  1.100 400     0.4 0.16 9 km 424.763  0.600 200 (30 km) 10 km 0.5 0.76 15 km

118.750  1.500 400     0.4 0.16 7 km 424.763  1.000 400     0.5 0.54 12 km

118.750  2.100 800     0.3 0.11 5 km 424.763  1.500 600     0.5 0.44 8 km

118.750  3.000 1000     0.2 0.10 3 km 424.763  4.000 1000     0.4 0.34 5 km

118.750  5.000 2000     0.1 0.07 surface

380.197  18.000 2000 1 pixel 12 km 1.0 0.72 6 km

424.763  4.000 1000 (10 km) 10 km 1.0 1.02 5 km

NET compliance code Compliant Nearly compliant Compliant on 2 x or 1 h Candidate to be dropped

Radiometric performance assessment for 15 min observing cycle

14

Azimuth Motor& Compensator

Elevation Motor& Compensator

Nodding / MorphingSubreflector

Thin Struts

Space Calibration Tube

BackupStructure

3” Thick Composite Reflector

50-430 GHz Feeds

Receiver Package

The GOMAS instrument with its 3-m antenna

15

S-Band Antenna for TT&C S-band antenna

for LRIT

   

NORTH DIRECTION

Space Calibration

Tube

NADIR

Solar Wing

3 -meterAntenna

Star Sensors

S-band Antennafor TT&C

The GOMASsatellite

•Mass: 860 kg (“dry": 430 kg)•Electrical power: 500 W•Volume: 3.0 x 3.0 x 3.0 m3

•Data rate: 128 kbps (S-band, compatible with MSG LRIT).

16

Team leader Institute Team leader Institute

Umberto AMATO CNR Istituto Applicazioni della Matematica, Napoli, Italy Paul MENZEL NOAA/NESDIS Office of Research and Application, Madison Wis., USA

John BATES NOAA/NESDIS, National Climatic Data Center, Asheville NC, USA Jungang MIAO Institute of Environmental Physics, University of Bremen, Germany

Wolfgang BENESCH Deutscher Wetterdienst, Offenbach, Germany Alberto MUGNAI CNR Istituto Scienze dell'Atmosfera e del Clima, Roma, Italy

Stefan BÜHLER Institute of Environmental Physics, University of Bremen, Germany Paolo PAGANO Servizio Meteorologico dell'Aeronautica, Roma, Italy

Massimo CAPALDO Servizio Meteorologico dell'Aeronautica, Roma, Italy Jean PAILLEUX Météo France, Toulouse, France

Marco CERVINO CNR Istituto Science dell'Atmosfera e del Clima, Bologna, Italy Juan PARDO Instituto de Estructura de la Materia, Madrid, Spain

Vincenzo CUOMO CNR Istituto Metodologie Avanzate di Analisi Ambientale, Potenza, Italy Federico PORCU' Department of Physics, University of Ferrara, Italy

Luigi De LEONIBUS Servizio Meteorologico dell'Aeronautica, Roma, Italy Catherine PRIGENT Dep.nt de Radioastronomie Millimetrique, Observatoire de Paris, France

Michel DESBOIS CNRS Laboratoire de Météorologie Dynamique, Palaiseau, France Franco PRODI CNR Istituto Science dell'Atmosfera e del Clima, Bologna, Italy

Stefano DIETRICH CNR Istituto Science dell'Atmosfera e del Clima, Roma, Italy Rolando RIZZI Department of Physics, University of Bologna, Italy

Frank EVANS University of Colorado, Atmospheric & Oceanic Sciences, Boulder Co., USA Guy ROCHARD MétéoFrance, Centre de Météorologie Spatiale, Lannion, France

Laurence EYMARD CNRS Centre études Environnement Terrestres et Planétaires, Vélizy, France Hans Peter ROESLI MétéoSuisse, Locarno-Monti, Switzerland

Albin GASIEWSKI NOAA Environmental Technology Laboratory, Boulder Co., USA Carmine SERIO Dip.nto Ingegneria e Fisica Ambiente, University of Basilicata, Potenza, Italy

Nils GUSTAFSSON Swedish Meteorological and Hydrological Institute, Norrköping, Sweden William SMITH NASA Langley Research Center, Hampton VA., USA

Georg HEYGSTER Institute of Environmental Physics, University of Bremen, Germany Antonio SPERANZA Hydrographic and Mareographic Service, Roma, Italy

Marian KLEIN NOAA Environmental Technology Laboratory, Boulder Co., USA David STAELIN MIT Research Laboratory of Electronics, Cambridge MA., USA

Klaus KÜNZI Institute of Environmental Physics, University of Bremen, Germany Alfonso SUTERA Department of Physics, University of Roma, Italy

Vincenzo LEVIZZANI CNR Istituto Science dell'Atmosfera e del Clima, Bologna, Italy Jung-Jung TSOU NASA Langley Research Center, Hampton VA., USA

Gian Luigi LIBERTI CNR Istituto Science dell'Atmosfera e del Clima, Bologna, Italy Chris VELDEN Cooperative Institute for Meteorological Satellite Studies, Madison Wis.,USA

Ernesto LOPEZ-BAEZA Dep.nt of Thermodynamics, Faculty of Physics, University of Valencia, Spain Guido VISCONTI Department of Physics, University of L'Aquila, Italy

List of Proponents of GOMAS (undertaking to implement the scientific programme)

P.I.: Bizzarro BIZZARRI, for the CNR Istituto Scienze dell'Atmosfera e del Clima, Roma, Italy

17

Conclusions 

• Strong requirements exist for frequent precipitation observation.

• From GEO, a new physical principle needs to be exploited.

• GOMAS is proposed as a demonstration mission.

• It would be a precursor for future operational applications.

• From the technical standpoint, and building on the studies conducted in the U.S. on GEM, it is believed that no enabling technology is currently missing.

• The GOMAS satellite could be developed in time for a launch in the 2007-2009 timeframe.

• It would provide simultaneous retrieval of: - temperature profile (x 30 km) - humidity profile (x 20 km) - cloud liquid/ice water total column and gross profile (x 20 km) - precipitation rate (x 10 km)

each 15 minutes ! over 1 / 12 of the disk covering sea and land !