Observations of Mediterranean Precipitating Systems using AMSU Beatriz FUNATSU 1, Chantal CLAUD 1 and Jean-Pierre CHABOUREAU 2 1 Laboratoire de Meteorologie Dynamique/IPSL, Palaiseau 2 Laboratoire d Aerologie/CNRS-UPS, Toulouse
MOTIVATION We aim at exploring the information provided by the Advanced Microwave Sounding Unit (AMSU) as an alternative to the use of reanalysis data to form a climatology of precipitating systems in the Mediterranean region, without relying on retrieval algorithms. WHY AMSU? Microwave radiation penetrates clouds, and is sensitive to atmospheric temperature and moisture, cloud liquid water, cloud ice water, rain and surface wind speeds (Kidder et al., 2000) AMSU has improved horizontal and spectral (hence vertical) resolutions compared to its predecessor (MSU) typical Med. cyclones and cloud systems are subsynoptic scale (Trigo et al. 1999, Chaboureau and Claud 2006), therefore features that were not captured using MSU may be detected using this dataset. Continuous coverage: AMSU instrument has been collecting data onboard NOAA satellites since 1998.
OBJECTIVES Upper level troughs are frequently observed upstream of areas of cloudiness/strong precipitation (e.g., Massacand et al. 1998, Chaboureau and Claud 2006). Use satellite data only to: 1.Identify upper level precursors of storms; 2.Detect and characterize storms; 3.Identify configurations favorable to the formation/occurrence of severe weather conditions. APPROACH Focus on selected case studies of extreme precipitation events, e.g.: Algiers [9-11 Nov 2001] 262 mm/24h in Algiers, 240 mm/24h in Balears; Israel [3-5 Dec 2001] 250 mm/24h in Zichron Yaakov; Gard Region [9 Sep 2002] 690 mm/24h in some stations; Other events: Antalia (2002), Rhône-Herault (2003), Herault-Nîmes (2005) In addition: November 2001
OUTLINE 1 - Description of AMSU 2 Identification of upper level precursors of storms 3 - Detect and characterize storms 4 - Application for a case study 5 - Application for November 2001.
1. AMSU AMSU-A: 48 km @ nadir AMSU-B: 16 km @ nadir Height (km) 3 4 5 H2 2O v channels Window channels information about the surface O 2 channels 5-8 sample T in the mid to upper troposphere H 2 O channels 3-5 information about water vapor content
1. AMSU AMSU-A: 48 km @ nadir AMSU-B: 16 km @ nadir Height (km) 3 4 5 H2 2O v channels AMSU-A BTs need correction for limb effect; BT s were adjusted using coefficients derived from RTTOV based on the case studies mentioned before
Algiers 2 Identification of upper level precursors of storms solid: PV @ 250 hpa dotted: mean SLP negative PV anom. positive PV anom. PV = 2PVU Upper level troughs have strongest PV anomaly signature between 200-300 hpa Nîmes-Marseille Israel Which channels would be able to capture this signal?
2 Identification of upper level precursors of storms AMSU-A ~200 hpa ~300 hpa Upper level troughs have strong(est) PV anomaly signature between 200-300 hpa Which channels would be able to capture this signal?
Algiers AMSU-A ch [7-5] Israel Conclusion: Ch. 8 performs better in identifying upper level troughs than does the difference between ch 7 and 5. Nîmes-Marseille 2 Identification of upper level precursors of storms AMSU-A ch 8
3 Detect and characterize storms We seek two thresholds: one that is able to identify precipitating systems (in general), and another that will indicate areas with deep convective clouds. Ch. 3 to 5 : H 2 O channels - scattering by icy particles - useful to identify cold clouds Ch. 1 and 2 : window channels - observe the surface
3 Detect and characterize storms: Comparison with TRMM data: Nîmes-Marseille 02 UTC 9 September 2002 AMSU-B03 AMSU-B3m4 Blue: AMSU-B Yellow-shaded: TRMM (accum. rain 0-3 UTC) AMSU-B3m5 AMSU-B4m5 AMSU-B ch 3 minus 5-8 K was found to be a robust threshold to detect areas that yield accumulated precipitation of 10 mm/3h
3 Detect and characterize storms: «Deep Convection Threshold» Analysis using TRMM Frequency distribution for Tropical Deep Convection Threshold (Hong et al., 2005) DCT = AMSU_B45 0 and AMSU_B35 0 and AMSU_B34 0 100 Cumulative relative fre equency 90 80 70 60 50 40 30 20 10 0 Case studies (20) Nov 2001 (150) >0 >=5 >=10 >=20 >=50 >=100 Accumulated precipitation (mm/3hr)
3 Detect and characterize storms: Comparison with Ground Data: Nîmes-Marseille 02UTC 9 September 2002 AMSU-B3m5 and DCT Thin black radar (precip intensity in mm/h at 02:30UTC) Black accum. precip between 2-3 UTC - station SQR (precip. accum. during 24h) Radar and accumulated precip. provided by Brice Boudevillain (LTHE) SQR data provided by METEO-France (Véronique Ducrocq)
4 - Application for case of Algiers (9-11 Nov 2001) Location of upper level trough Moderate precipitation Deep convective clouds Solid black : A08 Thick black : A08 > 221 K Light blue : B3m5-8 K Magenta : DCT
5 - Application for November 2001 Solid black : AMSU-A08 Thick black : AMSU-A08 > 221 K [upper level trough] Light blue : AMSU-B3m5-8 K [precipitation] Magenta : DCT [deep convection]
SUMMARY AMSU AMSU-A A ch. 8 works best to identify upper level troughs, while ch. 7 minus 5 provides information about its vertical penetration. Combination of ch. 3 to 5 of AMSU-B is able to detect moderate to heavy rainfall. In particular AMSU-B ch. 3 minus 5-8 K was found to be a robust threshold to detect moderate precipitation. DCT is found to detect precipitating areas that yield at least 20 mm accumulated in 3hr, in nearly 50% of the cases when the condition is satisfied. Ongoing work: use the above channels to establish a climatology of precipitating systems and their typology relative to the upper level situation.