1. III) CHARACTERISTICS OF THE ADDED CLUTTER RAIN MeteoSvizzera, 6605 Locarno, Switzerlandemail: Katja.Friedrich@meteoswiss.chhttp://www.meteoswiss.ch Simulation of Ground Clutter Influence on Polarimetric Rainfall Rate Katja Friedrich*, Urs Germann*, Gianmario Galli*, Pierre Tabary +, Jacques Parent du Chatelet + * MeteoSwiss, Locarno, Switzerland; + Météo France, Paris, France MOTIVATION How sensitive is the polarimetric rainfall rate estimation to ground clutter contamination when the clutter intensity is as large as or smaller than that of rain? We investigate the sensitivity of quantitative precipitation estimation during stratiform precipitation in the presence of strong and weak clutter contamination caused by obstacles such as mountains, buildings, trees, power lines. IDEA REALIZATION -- Take I/Q time series of clutter and rain measured by the Météo France C-band polarimetric Doppler radar located ~30 km southwest of Paris -- Create a mask representing the location of rain based on thresholds -- Select I/Q time series of point targets (Eiffeltower) and more complex clutter targets with different polarimetric characteristics -- Add continuously the signatures of along the range to observed I/Q time series of rain for an entire scan; repeat this by reducing the power of clutter signals step by step until its median in reflectivity is < 0.3 dBZ; determine the critical level of clutter signal for polarimetic rainfall rate estimation Distance / azimuthal range (km/°) Z h (dBZ) Z dr (dB)  hv  dp (°) Rain76.8 / 3.032.9 (32.9)0.40.998.3 B124.0 / 0.555.5 (67.9)4.90.9651.1 B224.0 / 3.049.2 (61.6)7.30.9656.8 B33.8 / 3.528.5 (56.8)-4.40.3043.4 B420.9 / 2.525.6 (39.2)-3.70.87123.4 B50.7 / 2.596.7 (33.8)0.50.94-5.9 B61.4 / 2.589.9 (33.1)-3.20.91132.8 B71.68 / 2.586.8 (31.3)-0.90.53-11.2 B81.92 / 2.583.9 (29.6)-1.20.3312.4 B923.8 / 2.529.9 (0.4)5.30.81126.1 IV) REFERENCE FIELDS + ADDED CLUTTER = CONCLUSIONS based on the 15-Nov-case & OUTLOOK I) REFERENCE FIELDS – STATIFORM PRECIPIATION ON 15 NOVEMBER 2005 AT 1456 UTC Z h = 30.26 dBZ Z dr = 0.37 dB  dp = 7.74°  hv = 0.98 R(Z h,Z dr ) = 3.1 mm h -1 R(Z h ) = 2.0 mm h -1 R(Z h ) = 0.0034 Z h 0.6024 R(Z h, Z dr ) = 0.0221 Z h 0.76 10 -0.33Zdr B1 B2 B3 B7 B4 B9 B8 B5 B6 Z dr (dB)R(Z h ) (mm h -1 ) b) -30 dBZ R(Z h Z dr ) (mm h -1 ) 2.) Reduced power of the clutter signal versus differences between (reference fields + added clutter) and reference fields for B1 – B9: a) -13 dBZ 1.) Reference fields (rain) versus reference fields+B1 (rain+clutter) when the median of [Z h (clutter) – Z h (rain)] is: 1a) -13 dBZ 1b) -30 dBZ Z h (dBZ)  dp (°) -13 dBZ -30 dBZ Median Z h (dBZ) Median Z dr (dB) Median  dp (°) Assuming theoretical accuracies as shown in (black, horizontal lines), the critical level of clutter contamination for polarimetric parameters (black, vertical lines) is achieved when the clutter signal is: ♪ 6 – 11 dBZ lower than that of rain for Z h ♪ 7 – 20 dBZ lower than that of rain for Z dr ♪10 – 26 dBZ lower than that of rain for  hv ♪ 6 – 23 dBZ lower than that of rain for  dp Variability of the critical level accuracy Assuming an theoretical accuracy in rain rate of ~10 %, the critical level for rainfall rate estimation is achieved when the clutter signal is: ♪7 – 13 dBZ lower than that of rain for R(Z h ) ♪6 – 24 dBZ lower than that of rain for R(Z h,Z dr ) Polarimetric variables and R(Z h,Z dr ) are more sensitive to small amounts of ground clutter but also have a larger variability of the critical level compared to Z h and R(Z h ) II) RAIN MASK 25 < Range < 100 km 15 < Z h < 50 dBZ Texture(Z dr ) < 1.50 dB 0 < Z dr < 4 dB  hv > 0.97  dp < 20° Beam shielding < 50 % 1. 1. 4. 3. 2. 2. 3. 4. 4.1. 3. Z h (dBZ)Z dr (dB)  dp (°)  hv R(Z h ) (mm h -1 ) R(Z h,Z dr ) (mm h -1 ) Ground clutter Rain Z h (rain+clutter) (dBZ) Z dr (rain+clutter) (dB) R(Z h ) (rain+clutter) (mm h -1 ) R(Z h,Z dr ) (rain+clutter) (mm h -1 )  dp (rain+clutter) (°)  dp (rain) (°) R(Z h,Z dr ) (rain) (mm h -1 ) R(Z h ) (rain) (mm h -1 )Z dr (rain) (dB)Z h (rain) (dBZ) 4.1.1. 4.1.2. 4.1.1. 4.1.2. 4.2. 4.2. Measure in-phase (I) & quadrature (Q) signals in clear air and rain Add I/Q time series measured in clutter to I/Q time series of rain Process data, calculate polarimetric variables (Z h, Z dr,  hv,  dp ), apply a quality control, estimate the rain rate R(Z h ), R(Z h, Z dr ) Compare results  0.1 dB 1 dBZ Median R(Z h ) (mm h -1 ) ~10 % (0.2 mm h -1 )  1° Median R(Z h Z dr ) (mm h -1 ) Median  hv 0.005 ~10 % (  0.3 mm h -1 ) These results need to be verified by more cases with different rain rates and drop size distributions and the accuracy of R(K dp ), R(K dp, Z dr ) needs to be included. 35.036.037.038.0 Azimuth (°) Z h (rain+clutter) – Z h (rain) (dBZ) Z dr (rain+clutter) – Z dr (rain)- (dB) R(Z h ) (rain+clutter) – R(Z h ) (rain) (mm h -1 ) R(Z h,Z dr ) (rain+clutter) – R(Z h,Z dr ) (rain) (mm h -1 )  dp (rain+clutter) –  dp (rain) (°)  hv (rain+clutter) –  hv (rain) (°) Z h (clutter) – Z h (rain) (dBZ) 3.