Real-time Estimation of Precipitation Using WSR-88D Weather Radars David R. Legates, Ph.D., C.C.M. Associate Professor and Director Center for Climatic Research University of Delaware Newark, Delaware 19716

1
Real-time Estimation of Precipitation
UsingWSR-88D Weather RadarsDavid R. Legates,
Ph.D., C.C.M.Associate Professor and
DirectorCenter for Climatic ResearchUniversity
of DelawareNewark, Delaware 19716
2
THREE TYPES OF ANALYSES
Climatological Precipitation Estimates Versus S
easonal Precipitation Totals Trends
Versus Real-Time Precipitation Estimates
3

• High Resolution Weather Data System
• Originally Sponsored by Duke Energy Corporation
• of Charlotte, North Carolina
• Initial Application
• Provide the front-end to Duke Energys River
  Management System of the Catawba River Basin for
  input to their Power Load Management System

4

• High Resolution Weather Data System
• Station Data Products
• Air Pressure S Gage Precipitation
• Air Temperature l Solar Radiation
• Dew Point Temperature l Wind Vector
• WSR-88D Radar Products
• Radar-Based Precipitation
• Composite Gage-Radar Precipitation
• Derived Products
• 12-Hr Precipitation S Precip. Difference Fields
• 24-Hr Precipitation S Storm Total Precipitation
• Relative Humidity l Apparent Temperature
• Evapotranspiration l Soil Moisture Content

5
(No Transcript)
6
(No Transcript)
7
(No Transcript)
8
(No Transcript)
9
(No Transcript)
10
(No Transcript)
11
(No Transcript)
12
(No Transcript)
13
(No Transcript)
14
(No Transcript)
15
(No Transcript)
16
(No Transcript)
17
(No Transcript)
18
(No Transcript)
19
(No Transcript)
20
(No Transcript)
21
National Weather Service WSR-88D Weather
Radars NEXRAD 10 cm wavelength Doppler-based
22
(No Transcript)
23
(No Transcript)
24
(No Transcript)
25
(No Transcript)
26
(No Transcript)
27
(No Transcript)
28
(No Transcript)
29
(No Transcript)
30
Reflectivity ZRainfall Rate Rwhere D
is the raindrop diameter, NB(D) and NG(D) are the
dropsize distributions at the height of the beam
and ground, respectively, and FT(D) is the
terminal fall velocity.
31

• WSR-88D Precipitation Processing
• Digital Precipitation Array (DPA)
• Precipitation Processing Algorithms
• Account for radar beam blockage
• Check for spurious noise and outliers
• Ground return/tilt test (0.5 versus 1.5 tilt
  angles)
• Construction of Hybrid Scan
• Precipitation Rate Algorithms
• Z-R relationship is applied -- usually Z 300
  R1.4
• Simple averaging from 2km to 4km resolution
• Time continuity checks
• Precipitation Accumulation Algorithms
• Scan and hourly accumulations
• Missing data and outliers check
• Precipitation Adjustment Algorithms
• NOT IMPLEMENTED

32

• Errors in Radar Precipitation Estimates
• Errors associated with reflectivity sign
  range
• Ground Clutter Contamination Yes
• Anomalous Propagation (Super-refractive
  conditions) No
• Partial Beam Filling Yes
• Wet Radome Attenuation No
• Attenuation by Oxygen, Water Vapor, Clouds,
  Rainfall Yes
• Incorrect Hardware Calibration / No
• Errors associated with the Z-R relationship
• Variations in Dropsize Distribution / No
• Hail, Mixed Precipitation, and Snow Events
  No
• Errors associated with effects below the radar
  beam
• Advection -- Strong Horizontal Winds / Yes
• Virga -- Evaporation of Falling Precipitation
  Yes
• Condensation/Coalescence Below Radar Beam
  Yes
• Vertical Motions -- Updrafts and Downdrafts
  / No
• WSR-88D system claims to specifically address
  these problems

33
Calibration, therefore, uses the WSR-88D radar
data for the spatial footprint of the storm and
adjusts the radar reflectivities using the gage
observations.
Gage Measurements Versus Radar Estimates

• Gage-Measured Precipitation
• Provides a good estimate of precipitation at a
  given point (when adjusted for gage measurement
  biases)
• Nearly all networks lack the gage densities
  needed to provide high-resolution estimates of
  storm-scale precipitation at an hourly time step

• Radar Precipitation Estimates
• Good spatial representation of precipitation is
  afforded by the DPAs 4km x 4km resolution
• Accuracy of precipitation estimates is very low
  due to errors associated with reflectivity,
  below-beam effects, and the Z-R relationship

34
(No Transcript)
35
(No Transcript)
36
KHGX Radar Calibration Oct 94
37
Radar Calibration Procedure

• Compute the DPA-composite reflectivity, Z,
  from Z 300 R 1.4 or Z 250 R 1.2
  or Z 200 R 1.6 (Standard) (Tropical)
  (Stratiform) where R is the precipitation
  estimate from the DPA. The appropriate equation
  is chosen from the Z-R relationship used by the
  NWS to derive the DPA.
• Then, compute the Composite Gage-Radar
  precipitation estimate, R, using
• Z a Rb Dc
• where D is the range from the radar and a, b,
  and c are calibration parameters. Parameters are
  fit using weighted least-squares logarithmic
  regression and observed reflectivity-gage pairs.

38
Hourly Pair Calibration (Legates, 2000) where
a, b, and c are constants and D is distance of
the reflectivity from the radar Calibration is
made using radar-gage pairs computed on an hourly
interval
39
(No Transcript)
40
(No Transcript)
41
Problems in Estimating Snowfall using Weather
Radar

• See previous discussion about rainfall
• Fall rate is smaller which accentuates the
  timing/advection problem
• Reflectivity varies considerably between liquid
  and solid hydrometeors
• Not all solid hydrometeors are equal!
• Very few real-time observations of solid
  precipitation exist!!

42
Limitations in theLegates (2000) Method

• Scatter is relatively small within storm events
  and increases as differing storm events are
  included
• Distance adjustment is not always significant
  owing to the different elevation angles chosen
• System limits pair generation to one pair per
  update cycle per gage
• Timing issues may exacerbate advection problems

43
A New Physically-Based Approach

• Differentiate between storm events and regions
  within storms
• Incorporate distance adjustments based on the
  selection of elevation angles with distance
• Enhance pair generation to use all radar updates
  and more frequent gage observations
• More stringent controls on timing issues to
  reduce advection problems
• Include physical effects on reflectivity biases

44
A New Physically-Based Approach

• Select pairs according to similar rainfall events
  using surface meteorological conditions
• Potential Temperature
• Equivalent Potential Temperature
• Air Temperature and Dew Point Range
• Wind shift
• Atmospheric Pressure

45
(No Transcript)
46
(No Transcript)
47
A New Physically-Based Approach

• where the integral holds over the radial from the
  radome to the cell
• Thus, a, b, c, d, and the function f must be
  estimated, as well as a new selection of pairs.

48
Key Issues/Gap/Challenges

• There is a definite need for real-time, high
  spatial resolution estimates of solid and mixed
  precipitation events
• Onset/duration of the event AND
• Semi-quantitative assessments of solid
  hydrometeor water equivalent may be all that is
  necessary
• Weather radar may be the best solution to this
  problem

49
Real-time Estimation of Precipitation
UsingWSR-88D Weather RadarsDavid R. Legates,
Ph.D., C.C.M.Associate Professor and
DirectorCenter for Climatic ResearchUniversity
of DelawareNewark, Delaware 19716