RadarDerived Precipitation

1
Radar-Derived Precipitation

• Deriving Precipitation Rates
• Radar Sampling Issues
• Validating Comparing Radar Estimates with Gauge
  Reports

RFC/HPC Hydromet 01-1 Matt Kelsch Wednesday, 15
November 2000 kelsch_at_ucar.edu
2
(No Transcript)
3
(No Transcript)
4
(No Transcript)
5
(No Transcript)
6
(No Transcript)
7
Z-R RelationshipsWSR-88D, Marshall-Palmer
(general), and Tropical
8
(No Transcript)
9
(No Transcript)
10

• Sampling Issues
• Radar domain cannot be sampled at consistent
  elevations, with consistent bin volumes, or for
  precipitation with similar stage of development
  or phase.
• Range degradation
• Low-level beam blocking
• Changes in precip phase have inconsistent
  effects--bright band, hail contamination
• These are not effectively corrected by changing
  Z-R coefficients

11
(No Transcript)
12
(No Transcript)
13
16 sep99 Storm Total Radar-derived Accumulation
from KRAX (Raleigh NC)
14
16 sep99 Storm Total Radar-derived Accumulation
from KAKQ (Wakefield VA)
15
Bright Band ?
16
(No Transcript)
17
(No Transcript)
18
(No Transcript)
19
Radar-Rain Gauge Comparisons

• Radar samples a volume of the atmosphere
• At discrete intervals
• Up to several thousands feet AGL
• Over a surface area which may exceed 1 mi2
• Rain gauges sample
• Continuously
• At the surface
• Over an area less than 1 ft2
• Accumulations are measurements with the error
  factors associated with the gauge type

20
(No Transcript)
21
(No Transcript)
22
(No Transcript)
23
(No Transcript)
24
0500 UTC 7 Aug 1999
0700 UTC 7 Aug 1999
25
(No Transcript)
26
1215 UTC 27 June 1995
1815 UTC 27 June 1995
27
1402 UTC 27 June 1995
1658 UTC 27 June 1995
28
Virginia Topography
Radar-derived accumulation 27 June 1995
29
Satellite IR image with 15-min lightning strikes
at 0330 UTC 29 July 1997
30
KFTG WSR-88D 0.5
WSR-88D 0.5o tilt for 0329 UTC 29 July 1997
31
Changing Z-R

• Will help when
• Consistently different average DSD (climate)
• Tropical versus mid-latitude (warm vs. cold
  process)
• Maritime versus continental
• Consistently different average DSD (season)
• Convective versus stratiform
• Is not the solution when
• Range degradation, overshooting low-levels
• Phase change hail, melting snow
• Snowfall

32
KRAX Storm Total 1159 UTC 6 Sep 96 Z300R1.4
33
KRAX Storm Total 1159 UTC 6 Sep 96 Z250R1.2
34
Radar-derived PrecipitationA Summary Of Major
Points

• Radar provides excellent storm-scale information
  about the spatial and temporal evolution of
  precipitation systems.
• Radar provides very valuable input as part of a
  comprehensive, multi-sensor precipitation system.
• Quantitative reliability issues are related to
  the fact that radar samples some volume at some
  elevation to estimate precipitation at the
  ground.
• Radar-derived precipitation is most reliably
  modeled for liquid hydrometeors hail and snow
  add complexity.
• The above two points are not effectively
  corrected by changing Z-R coefficients Z-R
  changes should be related to Drop Size
  Distribution knowledge.
• Radars and rain gauges do not measure equal
  samples
• Rain gauges do not provide a good representation
  of precipitation distribution, especially
  convective precip.
• The PPS algorithm has the versatility to evolve
  into a more comprehensive system, taking into
  account the ambient environment.