estimating raindrop size distributions and vertical motions using … · uazr observations at sgp...
TRANSCRIPT
![Page 1: Estimating Raindrop Size Distributions and Vertical Motions using … · UAZR observations at SGP Backup slides Huntsville, 2DVD: 21,000 minutes If we assume a gamma shape DSD, there](https://reader034.vdocuments.net/reader034/viewer/2022051808/600bb8d83ddeb84c4910c2f3/html5/thumbnails/1.jpg)
Estimating Raindrop Size Distributions and Vertical Motions using S-band & KAZR
Vertically Pointing Radars
Christopher R. Williams Cooperative Institute for Research in
Environmental Sciences (CIRES) In partnership with
NOAA Earth System Research Laboratory (ESRL)
Mid-latitude Continental Convective Cloud Experiment
(MC3E)
Support for this work: DOE ASR Grant: DE-SC0007080 NASA PMM Grant: NNX13AF89G
![Page 2: Estimating Raindrop Size Distributions and Vertical Motions using … · UAZR observations at SGP Backup slides Huntsville, 2DVD: 21,000 minutes If we assume a gamma shape DSD, there](https://reader034.vdocuments.net/reader034/viewer/2022051808/600bb8d83ddeb84c4910c2f3/html5/thumbnails/2.jpg)
Research Objectives • Goal
‒ Estimate profiles of vertical air motion and raindrop size distributions (DSDs) using two non-synchronous and non-beam matched vertically pointing profilers
• Approach ‒ Exploit the different backscattering signatures at Ka-
band (KAZR) and S-band • KAZR – non-Rayleigh (also, non-Bragg) • S-band – Rayleigh
• Key Results ‒ Air motion & DSD retrievals in rain ‒ Retrievals independent of radar calibration ‒ Nw and rain rate dependent on S-band calibration
![Page 3: Estimating Raindrop Size Distributions and Vertical Motions using … · UAZR observations at SGP Backup slides Huntsville, 2DVD: 21,000 minutes If we assume a gamma shape DSD, there](https://reader034.vdocuments.net/reader034/viewer/2022051808/600bb8d83ddeb84c4910c2f3/html5/thumbnails/3.jpg)
MC3E Profiling Radar Deployment
![Page 4: Estimating Raindrop Size Distributions and Vertical Motions using … · UAZR observations at SGP Backup slides Huntsville, 2DVD: 21,000 minutes If we assume a gamma shape DSD, there](https://reader034.vdocuments.net/reader034/viewer/2022051808/600bb8d83ddeb84c4910c2f3/html5/thumbnails/4.jpg)
Dual Frequency Air Motion Retrieval
Bragg Scattering
Rayleigh Scattering S-band Profiler Spectra
449 MHz Profiler Spectra
Air Motion Retrieval
![Page 5: Estimating Raindrop Size Distributions and Vertical Motions using … · UAZR observations at SGP Backup slides Huntsville, 2DVD: 21,000 minutes If we assume a gamma shape DSD, there](https://reader034.vdocuments.net/reader034/viewer/2022051808/600bb8d83ddeb84c4910c2f3/html5/thumbnails/5.jpg)
449 MHz 20 May 2011 12:27:33 UTC
Bragg Scattering
Rayleigh Scattering
![Page 6: Estimating Raindrop Size Distributions and Vertical Motions using … · UAZR observations at SGP Backup slides Huntsville, 2DVD: 21,000 minutes If we assume a gamma shape DSD, there](https://reader034.vdocuments.net/reader034/viewer/2022051808/600bb8d83ddeb84c4910c2f3/html5/thumbnails/6.jpg)
1-Minute Dwell – Pros & Cons Profiles from KAZR and S-band radars are:
• not simultaneous • not matched beams
Average spectra for 1-minute
• Represent larger spatial domain • Cannot represent the turbulence within pulse
volume Number of profiles per 1-minute dwell:
• KAZR – 17 profiles per minute • S-band – 7 profiles per minute
![Page 7: Estimating Raindrop Size Distributions and Vertical Motions using … · UAZR observations at SGP Backup slides Huntsville, 2DVD: 21,000 minutes If we assume a gamma shape DSD, there](https://reader034.vdocuments.net/reader034/viewer/2022051808/600bb8d83ddeb84c4910c2f3/html5/thumbnails/7.jpg)
Observed Doppler Velocity Spectra
![Page 8: Estimating Raindrop Size Distributions and Vertical Motions using … · UAZR observations at SGP Backup slides Huntsville, 2DVD: 21,000 minutes If we assume a gamma shape DSD, there](https://reader034.vdocuments.net/reader034/viewer/2022051808/600bb8d83ddeb84c4910c2f3/html5/thumbnails/8.jpg)
Observed Doppler Velocity Spectra
Velocity Difference is due to DSD shape. (independent of air motion)
Shift in mean velocity by both radars is due to air motion. (independent of DSD shape)
![Page 9: Estimating Raindrop Size Distributions and Vertical Motions using … · UAZR observations at SGP Backup slides Huntsville, 2DVD: 21,000 minutes If we assume a gamma shape DSD, there](https://reader034.vdocuments.net/reader034/viewer/2022051808/600bb8d83ddeb84c4910c2f3/html5/thumbnails/9.jpg)
Mean Doppler Velocity DSD (scaled PDF): Radar Weighted Fall Velocity: 𝑆ℎ𝑦𝑦𝑦𝑦 𝑣;𝑁𝑤 ,𝐷𝑚, 𝜇 = 𝑁𝑤𝑓(𝐷;𝐷𝑚, 𝜇)𝜎𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑦(𝐷)𝑔 𝐷, 𝑣
Mean Hydrometer Fall Velocity:
𝑉ℎ𝑦𝑦𝑦𝑦 𝐷𝑚, 𝜇 =∫ 𝑆ℎ𝑦𝑦𝑦𝑦 𝑣;𝐷𝑚, 𝜇 𝑣𝑣𝑣∫ 𝑆ℎ𝑦𝑦𝑦𝑦 𝑣;𝐷𝑚, 𝜇 𝑣𝑣
Measured Velocity (Doppler Velocity):
‘Intensity’ ‘quasi PDF’
𝑉𝑚𝑏𝑏𝑏𝑚𝑦𝑏𝑦 =∫ 𝑆 𝑣 𝑣𝑣𝑣𝑣𝑚𝑚𝑚𝑣𝑚𝑚𝑚
∫ 𝑆 𝑣 𝑣𝑣𝑣𝑚𝑚𝑚𝑣𝑚𝑚𝑚
= 𝑉ℎ𝑦𝑦𝑦𝑦 𝐷𝑚, 𝜇 −𝑊
𝑁 𝐷;𝑁𝑤 ,𝐷𝑚, 𝜇 = 𝑁𝑤 𝑓(𝐷;𝐷𝑚, 𝜇)
Diameter-to-fall conversion
![Page 10: Estimating Raindrop Size Distributions and Vertical Motions using … · UAZR observations at SGP Backup slides Huntsville, 2DVD: 21,000 minutes If we assume a gamma shape DSD, there](https://reader034.vdocuments.net/reader034/viewer/2022051808/600bb8d83ddeb84c4910c2f3/html5/thumbnails/10.jpg)
Difference in Doppler Velocity (DDV) Mean Doppler Velocities: Difference in Doppler Velocities (DDV): Bottom line: Difference in Doppler velocity (DDV) is independent of air motion and only dependent on the shape of the DSD.
𝑉𝑚𝑏𝑏𝑏𝑚𝑦𝑏𝑦𝑆𝑏𝑏𝑆𝑦 = 𝑉ℎ𝑦𝑦𝑦𝑦𝑆𝑏𝑏𝑆𝑦 𝐷𝑚, 𝜇 −𝑊
𝑉𝑚𝑏𝑏𝑏𝑚𝑦𝑏𝑦𝐾𝐾𝐾𝐾 = 𝑉ℎ𝑦𝑦𝑦𝑦𝐾𝐾𝐾𝐾 𝐷𝑚, 𝜇 −𝑊
𝐷𝐷𝑉 = 𝑉𝑚𝑏𝑏𝑏𝑚𝑦𝑏𝑦𝑆𝑏𝑏𝑆𝑦 − 𝑉𝑚𝑏𝑏𝑏𝑚𝑦𝑏𝑦
𝐾𝐾𝐾𝐾
𝐷𝐷𝑉 = 𝑉𝐻𝑦𝑦𝑦𝑦𝑆𝑏𝑏𝑆𝑦 𝐷𝑚, 𝜇 − 𝑉𝐻𝑦𝑦𝑦𝑦𝐾𝐾𝐾𝐾 (𝐷𝑚, 𝜇)
![Page 11: Estimating Raindrop Size Distributions and Vertical Motions using … · UAZR observations at SGP Backup slides Huntsville, 2DVD: 21,000 minutes If we assume a gamma shape DSD, there](https://reader034.vdocuments.net/reader034/viewer/2022051808/600bb8d83ddeb84c4910c2f3/html5/thumbnails/11.jpg)
Simulated Doppler Spectra
![Page 12: Estimating Raindrop Size Distributions and Vertical Motions using … · UAZR observations at SGP Backup slides Huntsville, 2DVD: 21,000 minutes If we assume a gamma shape DSD, there](https://reader034.vdocuments.net/reader034/viewer/2022051808/600bb8d83ddeb84c4910c2f3/html5/thumbnails/12.jpg)
DFR and DDV vs. Mean Diameter
![Page 13: Estimating Raindrop Size Distributions and Vertical Motions using … · UAZR observations at SGP Backup slides Huntsville, 2DVD: 21,000 minutes If we assume a gamma shape DSD, there](https://reader034.vdocuments.net/reader034/viewer/2022051808/600bb8d83ddeb84c4910c2f3/html5/thumbnails/13.jpg)
DFR & DDV with Norm. Breadth
𝜎′𝑚=0.24 𝜎′𝑚=0.24 𝜎′𝑚=0.24
𝜎′𝑚=0.24 𝜎′𝑚=0.24 𝜎′𝑚=0.24
![Page 14: Estimating Raindrop Size Distributions and Vertical Motions using … · UAZR observations at SGP Backup slides Huntsville, 2DVD: 21,000 minutes If we assume a gamma shape DSD, there](https://reader034.vdocuments.net/reader034/viewer/2022051808/600bb8d83ddeb84c4910c2f3/html5/thumbnails/14.jpg)
Retrieval Sequence Two Measurements:
KAZR Doppler Velocity Sband Doppler Velocity
Three Unknowns: Air motion (W) Mean Diameter (𝐷𝑚) Breadth Parameter (𝜎′𝑚)
Assumptions: Zero air motion at lowest range gate Breadth Parameter constant with ht
![Page 15: Estimating Raindrop Size Distributions and Vertical Motions using … · UAZR observations at SGP Backup slides Huntsville, 2DVD: 21,000 minutes If we assume a gamma shape DSD, there](https://reader034.vdocuments.net/reader034/viewer/2022051808/600bb8d83ddeb84c4910c2f3/html5/thumbnails/15.jpg)
Retrieval Sequence Two Measurements:
KAZR Doppler Velocity Sband Doppler Velocity
Three Unknowns: Air motion (W) Mean Diameter (𝐷𝑚) Breadth Parameter (𝜎′𝑚)
Assumptions: Zero air motion at lowest range gate Breadth Parameter constant with ht
𝑉𝑚𝑏𝑏𝑏𝑚𝑦𝑏𝑦𝑆𝑏𝑏𝑆𝑦 = 8.1 m/s
DDV = 0.79 m/s
![Page 16: Estimating Raindrop Size Distributions and Vertical Motions using … · UAZR observations at SGP Backup slides Huntsville, 2DVD: 21,000 minutes If we assume a gamma shape DSD, there](https://reader034.vdocuments.net/reader034/viewer/2022051808/600bb8d83ddeb84c4910c2f3/html5/thumbnails/16.jpg)
Step #1: Lowest range gate (160 m)
𝑉𝑚𝑏𝑏𝑏𝑚𝑦𝑏𝑦𝑆𝑏𝑏𝑆𝑦 = 8.16 m/s
DDV = 0.79 m/s
![Page 17: Estimating Raindrop Size Distributions and Vertical Motions using … · UAZR observations at SGP Backup slides Huntsville, 2DVD: 21,000 minutes If we assume a gamma shape DSD, there](https://reader034.vdocuments.net/reader034/viewer/2022051808/600bb8d83ddeb84c4910c2f3/html5/thumbnails/17.jpg)
Step #1: Lowest range gate (160 m)
𝑉𝑚𝑏𝑏𝑏𝑚𝑦𝑏𝑦𝑆𝑏𝑏𝑆𝑦 = 8.16 m/s
DDV = 0.79 m/s
𝜎′𝑚=0.36
𝜎′𝑚=0.24
𝜎′𝑚=0.26 (Best match)
![Page 18: Estimating Raindrop Size Distributions and Vertical Motions using … · UAZR observations at SGP Backup slides Huntsville, 2DVD: 21,000 minutes If we assume a gamma shape DSD, there](https://reader034.vdocuments.net/reader034/viewer/2022051808/600bb8d83ddeb84c4910c2f3/html5/thumbnails/18.jpg)
Retrieval Sequence Two Measurements:
KAZR Doppler Velocity Sband Doppler Velocity
Three Unknowns: Air motion (W) Mean Diameter (𝐷𝑚) Breadth Parameter (𝜎′𝑚)
Assumptions: Zero air motion at lowest range gate Breadth Parameter constant with ht
𝑉𝑚𝑏𝑏𝑏𝑚𝑦𝑏𝑦𝑆𝑏𝑏𝑆𝑦 = 7.3 m/s
DDV = 0.56 m/s
![Page 19: Estimating Raindrop Size Distributions and Vertical Motions using … · UAZR observations at SGP Backup slides Huntsville, 2DVD: 21,000 minutes If we assume a gamma shape DSD, there](https://reader034.vdocuments.net/reader034/viewer/2022051808/600bb8d83ddeb84c4910c2f3/html5/thumbnails/19.jpg)
Step #1: Lowest range gate
𝑉𝑚𝑏𝑏𝑏𝑚𝑦𝑏𝑦𝑆𝑏𝑏𝑆𝑦 = 7.30 m/s
DDV = 0.56 m/s 𝜎′𝑚=0.26 (Best match)
Difference is air motion
![Page 20: Estimating Raindrop Size Distributions and Vertical Motions using … · UAZR observations at SGP Backup slides Huntsville, 2DVD: 21,000 minutes If we assume a gamma shape DSD, there](https://reader034.vdocuments.net/reader034/viewer/2022051808/600bb8d83ddeb84c4910c2f3/html5/thumbnails/20.jpg)
Retrieval at 12:27 UTC
![Page 21: Estimating Raindrop Size Distributions and Vertical Motions using … · UAZR observations at SGP Backup slides Huntsville, 2DVD: 21,000 minutes If we assume a gamma shape DSD, there](https://reader034.vdocuments.net/reader034/viewer/2022051808/600bb8d83ddeb84c4910c2f3/html5/thumbnails/21.jpg)
Air motions: 12:25 to 12:30 UTC
![Page 22: Estimating Raindrop Size Distributions and Vertical Motions using … · UAZR observations at SGP Backup slides Huntsville, 2DVD: 21,000 minutes If we assume a gamma shape DSD, there](https://reader034.vdocuments.net/reader034/viewer/2022051808/600bb8d83ddeb84c4910c2f3/html5/thumbnails/22.jpg)
Time-height Cross-Section Z
![Page 23: Estimating Raindrop Size Distributions and Vertical Motions using … · UAZR observations at SGP Backup slides Huntsville, 2DVD: 21,000 minutes If we assume a gamma shape DSD, there](https://reader034.vdocuments.net/reader034/viewer/2022051808/600bb8d83ddeb84c4910c2f3/html5/thumbnails/23.jpg)
Spectra at 10:31 UTC
![Page 24: Estimating Raindrop Size Distributions and Vertical Motions using … · UAZR observations at SGP Backup slides Huntsville, 2DVD: 21,000 minutes If we assume a gamma shape DSD, there](https://reader034.vdocuments.net/reader034/viewer/2022051808/600bb8d83ddeb84c4910c2f3/html5/thumbnails/24.jpg)
Convective profile, 10:31:00 UTC
![Page 25: Estimating Raindrop Size Distributions and Vertical Motions using … · UAZR observations at SGP Backup slides Huntsville, 2DVD: 21,000 minutes If we assume a gamma shape DSD, there](https://reader034.vdocuments.net/reader034/viewer/2022051808/600bb8d83ddeb84c4910c2f3/html5/thumbnails/25.jpg)
Next Steps 1. Convective Rain in rain portion only
Use profiles every 10 seconds 2. UAZR instead of S-band?
FASTER analysis suggests UAZR & S-band moments are very similar (Scott G. & Tami)
3. Produce a PI Product using 2 years of KAZR / UAZR observations at SGP
![Page 26: Estimating Raindrop Size Distributions and Vertical Motions using … · UAZR observations at SGP Backup slides Huntsville, 2DVD: 21,000 minutes If we assume a gamma shape DSD, there](https://reader034.vdocuments.net/reader034/viewer/2022051808/600bb8d83ddeb84c4910c2f3/html5/thumbnails/26.jpg)
Backup slides
![Page 27: Estimating Raindrop Size Distributions and Vertical Motions using … · UAZR observations at SGP Backup slides Huntsville, 2DVD: 21,000 minutes If we assume a gamma shape DSD, there](https://reader034.vdocuments.net/reader034/viewer/2022051808/600bb8d83ddeb84c4910c2f3/html5/thumbnails/27.jpg)
Huntsville, 2DVD: 21,000 minutes If we assume a gamma shape DSD, there is a relationship between 𝜎𝑚 − 𝐷𝑚 − 𝜇 (Assume the 𝐷𝑚𝑏𝑚 = ∞)
𝜎𝑚2 =𝐷𝑚2
𝜇 + 4
Power-law fit
Count in dB -3 dB = ½ cnt
𝜇 = 10
𝜇 = 5 𝜇 = 0
1. Can estimate 𝜎𝑚 from 𝐷𝑚 and 𝜇
2. Can estimate 𝜇 from 𝐷𝑚 and 𝜎𝑚 𝜇 =𝐷𝑚2
𝜎𝑚2− 4
Williams et al. 2014, JAMC
![Page 28: Estimating Raindrop Size Distributions and Vertical Motions using … · UAZR observations at SGP Backup slides Huntsville, 2DVD: 21,000 minutes If we assume a gamma shape DSD, there](https://reader034.vdocuments.net/reader034/viewer/2022051808/600bb8d83ddeb84c4910c2f3/html5/thumbnails/28.jpg)
Huntsville 2DVD: 21,000 minutes
Calculated 𝜎′𝑚 vs. 𝐷𝑚
74% of observations are within +/- 1 STD (a normal distribution would have 68%)
Calculated 𝜇 vs. 𝐷𝑚 (assume a gamma DSD) Observed 𝜎𝑚 vs. 𝐷𝑚
Normalized PDF of 𝜎′𝑚
Narrower
Broader
Williams et al. 2014, JAMC
Key attribute: 𝜎′𝑚 and 𝐷𝑚 are uncorrelated
![Page 29: Estimating Raindrop Size Distributions and Vertical Motions using … · UAZR observations at SGP Backup slides Huntsville, 2DVD: 21,000 minutes If we assume a gamma shape DSD, there](https://reader034.vdocuments.net/reader034/viewer/2022051808/600bb8d83ddeb84c4910c2f3/html5/thumbnails/29.jpg)
Difference in Doppler Velocity (DDV) Mean Doppler Velocities: Difference in Doppler Velocities (DDV): Difference in Doppler velocity (DDV) is independent of air motion and only dependent on the shape of the DSD.
𝑉𝑚𝑏𝑏𝑏𝑚𝑦𝑏𝑦𝑆𝑏𝑏𝑆𝑦 = 𝑉ℎ𝑦𝑦𝑦𝑦𝑆𝑏𝑏𝑆𝑦 𝐷𝑚,𝜎′𝑚 −𝑊
𝑉𝑚𝑏𝑏𝑏𝑚𝑦𝑏𝑦𝐾𝐾𝐾𝐾 = 𝑉ℎ𝑦𝑦𝑦𝑦𝐾𝐾𝐾𝐾 𝐷𝑚,𝜎′𝑚 −𝑊
𝐷𝐷𝑉 = 𝑉𝑚𝑏𝑏𝑏𝑚𝑦𝑏𝑦𝑆𝑏𝑏𝑆𝑦 − 𝑉𝑚𝑏𝑏𝑏𝑚𝑦𝑏𝑦
𝐾𝐾𝐾𝐾
𝐷𝐷𝑉 = 𝑉𝐻𝑦𝑦𝑦𝑦𝑆𝑏𝑏𝑆𝑦 𝐷𝑚,𝜎′𝑚 − 𝑉𝐻𝑦𝑦𝑦𝑦𝐾𝐾𝐾𝐾 (𝐷𝑚,𝜎′𝑚)