Particle Size Distribution (PSD) Working Group Update
and
Using GV Observations to Investigate NUBF Estimates
at DPR Footprints
Christopher R. Williams
Cooperative Institute for Research in
Environmental Sciences (CIRES)
University of Colorado Boulder
Particle Size Distribution (PSD) NUBF sub-group:
Walt Petersen (lead), Liang Liao, Robert Meneghini, Pierre Kirstetter,
Simone Tanelli, Ali Tokay, Christopher Williams, and David Wolff
This work supported by the
NASA Precipitation Measurement Mission (PMM)
Grant NNX10AM54G
NASA PMM DSD Working Group:
Bridging Ground Validation (GV) and Algorithms
PSD
Working
Group
GV Data
Ground
Validation
Particle Size Distribution (PSD)
(August 2014)
Algorithm
Development
Future Field
Campaigns
PSD WG General Objective: Use Ground
Validation (GV) data to support, or justify,
assumptions used in satellite retrieval algorithms.
Today’s Outline:
1) Accomplishments of DSD Working Group
2) Describe 5 PSD sub-groups & their focus to improve retrieval algorithms
3) Highlight activities in Non-Uniform Beam Filling (NUBF) sub-group
PSD Working Group Monthly phone calls: 3rd Thursday @ 11 AM Eastern (US).
Accomplishments of DSD WG
Normalized breadth – independent of Dm
Broad
Narrow
Gamma shape parameter
Mass spectrum breadth
Using surface 2DVD
observations,
calculated joint PDFs
of DSD parameters vs
Dm.
Provides expected
value and range of
DSD parameters for
probabilistic retrieval
algorithms.
Gamma shape & slope parameters
Williams, Bringi, Carey, Chandra, Gatlin, Haddad, Meneghini, Munchak,
Nesbitt, Petersen, Tanelli, Tokay, Wilson, and Wolff 2013: Adaptive Raindrop
Size Distribution Constraint for Probabilistic Rainfall Retrieval Algorithms,
submitted to J. Appl. Meteor. Climatol.
Radar Integral Tables
Adaptive power-law constraint being tested in combined algorithm.
From Joe Munchak
DFR = Dual Frequency Ratio
DWR = Dual Wavelength Ratio
1.5
𝜎𝑚 = 𝑎𝜎𝑦 𝐷𝑚
1
−4
𝑐𝑜𝑛𝑠𝑡𝑟𝑎𝑖𝑛𝑡 𝜇 = 2
𝑎𝜎𝑦 𝐷𝑚
𝑐𝑜𝑛𝑠𝑡𝑟𝑎𝑖𝑛𝑡
𝑎𝜎𝑦 ranged from 0.11 to 0.47
(very narrow to very broad)
(Limits of integration = 3*Dm)
(Spherical drops)
Broad
Black line: mu = 2
Narrow
Radiometer Integral Tables
Adaptive power-law constraint being tested in combined algorithm.
From Joe Munchak
Parameters needed for radiative
transfer calculations are also
integrated over the same DSD with
the same code to ensure consistency
between radar and radiometer
algorithm modules.
Narrow (𝑎𝜎𝑦 = 0.11)
Broad (𝑎𝜎𝑦 = 0.47)
Improved Working Group Focus & Leadership
• To include solid precipitation, change name of
working group:
– Particle Size Distribution (PSD) Working Group
• 25 members & 30 associate members
• Improved Leadership:
– Christopher Williams – DSD parameter
relationships and column observations
– Steve Nesbitt – building the column and moving
to aircraft observations
– Joe Munchak – algorithm prospective and issues
with ice & frozen PSDs
PSD Sub-Groups (Defined August 2014)
Themes for NASA PMM PSD Working Group
• Light & Heavy Rain Estimation
• Melting Layer Characteristics
• Non-Uniform Beam Filling (NUBF)
– PIA(Ka) / PIA(Ku) Relationship
– Chandra
– Ben Johnson
– Walt Petersen
– Christopher Williams
• Solid Precipitation
– Steve Nesbitt
• DSD / PSD Parameter Relationships – Brenda Dolan
Impacts of NUBF on DPR Retrieval Algorithms
Definitions
• Non-Uniform Beam Filling (NUBF)
– Non-homogenous rain in IFOV
•
NUBF affects not only measured Z within radar resolution volume,
but affects signal attenuation to that resolution volume
Impacts
• Severity of NUBF impact depends on the specific method in
question & whether it is a single- or dual-frequency retrieval
– NUBF probably affects the Ka-band more strongly than the Ku-band
– Methods that use the SRT will be affected in a different way than nonSRT methods
Background (Theory)
• Toshio Iguchi’s NUBF correction theory (2009) is rigorous & elegant
• Assumptions in NUBF theory:
1.
2.
Assumes k and Z are constant with range (maybe valid at nadir, but
not off-nadir)
Critical parameter is coefficient of variation in k: CV(k) = std(k) /
Expected value(k)
How can GV Observations help Evaluate NUBF?
Can GV Observations support or improve upon Iguchi’s NUBF
correction theory?
Topics discussed in NUBF sub-group:
• PIA ratio: PIA(Ka)/PIA(Ku)
– PSD variability (homogenous rain)
– NUBF variability
•
Quantify sub-grid DPR rain variability using GV radars (network
radars)
– Large datasets to get good statistics of different rain regimes
•
3D structure
– Horizontal structure
• Correlation length across multiple FOVs
– Vertical structure
• Relationships between horizontal & vertical structures
– View angle dependence (slant view through vertical structures)
•
Analyses need to include DSD parameters to get attenuation
estimates at Ku and Ka
Rain-free Surface Reference
Aj
Z m,ij
Surface Return
Zˆ m,i
Aj
Large data base
to get statistics across
different rain regimes
Relative NUBF = std(R) / Rref
Pierre Kirstetter et al. (QJRMS 2014)
WFF NPOL Moderate Convection and GPM DPR
Examine roles of beam geometry, filling, microphysical variability on products
B
B
A
A
A
NPOL, KDOX and
DPR well- matched
in ice @ (7-10km)
Differences in strong
convective cells
where NPOL
reveals that small
hail, graupel and
large drops exist
B
NPOL Observations (Wallops Island) from 07-Sept-2014
High resolution data to investigate horizontal & vertical structure
RHI scan
5 x 0.5 km grid boxes
Mean(Z)
Raw obs.
Std(Z) vs.
mean(Z)
Std(Z)
Combining Ku-PR NS, Ku-PR MS
and Ka-PR HS
Simone Tanelli & Steve Durden
International Partnerships
Themes for NASA PMM PSD Working Group
• Light & Heavy Rain Estimation
• Melting Layer Characteristics
• Non-Uniform Beam Filling (NUBF)
– PIA(Ka) / PIA(Ku) Relationship
– Chandra
– Ben Johnson
– Walt Petersen
– Christopher Williams
• Solid Precipitation
– Steve Nesbitt
• DSD / PSD Parameter Relationships – Brenda Dolan
Solutions Require International Partnerships
• How can PSD WG help international partners?
• How can International partners help the PSD WG?