Drop size distribution from WXT weather transmitter? Heikki Pohjola, Application specialist Thanks to Elena
Weather Transmitter acoustic rain sensor Ultrasonic wind sensor CPU board PTU module screw terminal ©Vaisala | date | Ref. code | Page 2
The Equations. Output of radar systems is expressed in Z or d. BZ. Z is reflectivity factor, Z = Ni. Di 6 Ni number of particles with diameter Di per unit volume of the atmosphere. Z varies from 0. 0000001 to 10000000 so we use instead logarithmic d. BZ In Hydrology, this rainfall intensity is then estimated from Z = a. Rb where a and b are empirical parameters e. g. Z = 200 R 1. 6 where R = rainfall rate in mm/h ©Vaisala | date | Ref. code | Page 3
WXT_Z Raincap sensor of WXT works as a disdrometer, and measures N and D in time unit. Thus, we can calculate reflectivity on ground Zg = Ni. Di 6 and avoid all fuzzy Z/R conversions ©Vaisala | date | Ref. code | Page 4
DSD field test measurements in Vantaa, July 2006 Drop size distribution varies in time and place two sets during the same day, same site ©Vaisala | date | Ref. code | Page 5
The Danger • Changes in time in one place are similar to changes in place on one moment • Thus, using one observation for interpreting the entire image is no good (Not the same day, just a scematic image) ©Vaisala | date | Ref. code | Page 6
Application Set one or more WXTs to blocked sectors of radar Combine d. BZ values for each image (every 2, 5, 6, 10 or 15 minutes) Metman 3 and IRIS have tools for this using Barnes method for radar+gauge combination (see IRIS hydromet option) (Barnes, S. L. , 1964: A technique for maximising details in numerical weather map analysis. Jour_. App. I_Meteor. , _3, 396 -409) ©Vaisala | date | Ref. code | Page 7 A very bad case of beam blocking in mountaineus environment
Other applications • Gapfilling between radars • Ground reference for vertical profile of reflectivity ©Vaisala | date | Ref. code | Page 8
Data • Drop size distribution data from WXT, disdrometer (RD-69) and POSS from Järvenpää • WXT: 21. 8. -> • RD-69: All Testbed campaings • POSS: All Testbed campaings ©Vaisala | date | Ref. code | Page 9
Is WXT´s DSD data accurate enough? • Comparison between disdrometer, WXT and POSS data. (In WXT specs accumation accuracy 5 %. ) In future: • If results are promising: • Weather radar applications like WXT-beam blockage correction, -cap filling, -adjustment, . . . ©Vaisala | date | Ref. code | Page 10
Back up slides ©Vaisala | date | Ref. code | Page 11
FYI: Radar equation P rr 2 LC|K|2 =Z Where Pr is the power in watts – this is what radar technically measure L is attenuation and other losses C is radar constant (set in calibration) |K|2 is the dielectric factor (roughly, ice or water) r is distance from radar ©Vaisala | date | Ref. code | Page 12
The Opportunity Global distribution of DSD is badly known (as disdrometers are expensive and thus rare) Climatological research, especially global precipitation projects such as GPM could benefit of a bunch of reasonably priced identical disdrometers at each ground validation site. The Global Precipitation Measurement (GPM) Program Objectives: • obtain increased understanding of clouds and rainfall processes • make frequent rainfall measurements on a global basis. Continuation of the partnership of Tropical Rainfall Measuring Mission (TRMM). Partners: NASA, ESA, Japanese Aviation Exploration Agency ©Vaisala | date | Ref. code | Page 13
Attenuation When there is precipitation between the radar and the target, some microwave power is lost. This is called attenuation, and it is a problem in C-band X-band. Dual polarization has solutions. On single polarization radars, the attenuation correction is in form of where C and E are empirical parameters, which depend on drop size distribution. Research of this could be interesting (in one of the Tekes projects ? ) as in Kumpula we have the independent measure of real attenuation (from dual pol). ©Vaisala | date | Ref. code | Page 14
Attenuation is a problem in C-band; dual polarization shows maginitude and promises a fix correction scheme Corrected ©Vaisala | date | Ref. code | Page 15 Not corrected
Validation with gauges (1 h accumulation) ©Vaisala | date | Ref. code | Page 16
Note on d. BZ-based d. BZ attenuation correction This is notorious, as any errors cumulate and the correction can explode with range. However, in single pol radar it is the best we can get. Sigmet colution against exploding correction has been set of two thresholds as below. I do admit dual pol solutions are better. But, we have even single pol radars in the product family. ©Vaisala | date | Ref. code | Page 17
Скачать презентацию Drop size distribution from WXT weather transmitter Heikki
16cf8ef71aed06fb3b46a446b9fe945e.ppt