Quantitative Precipitation Forecasting at the Hydrometeorological Prediction Center HPC

1
Quantitative Precipitation Forecasting at the
Hydrometeorological Prediction Center
(HPC) www.hpc.ncep.noaa.gov
Dan Petersen HPC Forecast Operations
Branch Dan.Petersen_at_noaa.gov [phone]
2
Quantitative Precipitation Forecasting at the
Hydrometeorological Prediction Center (HPC)Goals
of Presentation

• Short Range QPF Methods
• Short Range QPF Case Study
• Verification

3
Composing a QPF

• Short range ( lt12 hours )
• Forecast composed by blending the latest radar
  and satellite data with an analysis of
  Moisture/Lift/Instability and model output
• Long range ( gt12 hours )
• Forecast increasingly relies on model output of
  QPF, Moisture/Lift/Instability
• Adjustments are made for known model biases and
  latest model trends/verification/comparisons
  (including ensembles)

4
Composing a QPF ( lt12 hours)

• Radar
• Looping can show areas of training and
  propagation
• Review radar-estimated amounts-Be wary of beam
  blocking, bright bands, overshooting tops
  attenuation
• Compare observations to estimates (Z R
  relationship impact)
• Satellite
• Rainfall estimates from NESDIS/Satellite
  Analysis Branch
• Looping images can show areas of
  training/development
• Derived Precipitable Water, Lifted Indices,
  soundings, etc.

5
OH Valley Case Study-Using Models/Radar/Satellite
to Compose QPF GFS 18z-00z QPF June 14 2005 from
12z Run
6
OH Valley Case Study-Using Models/Radar/Satellite
to Compose QPFNAM 18z-00z QPF June 14 2005 from
12z Run
7
OH Valley Case Study-Using Models/Radar/Satellite
to Compose QPFHPC Forecast qpf 18z-00z QPF
Jun14-15 2005
8
OH Valley Case Study-Using Models/Radar/Satellite
to Compose QPFNAM Forecast CAPE/CIN 18z June14
2005
9
OH Valley Case Study-Using Models/Radar/Satellite
to Compose QPFNAM Forecast Precipitable Water
18z June14 2005
10
OH Valley Case Study-Using Models/Radar/Satellite
to Compose QPFNAM Forecast Best Lifted Indices
18z June14 2005
11
OH Valley Case Study-Using Models/Radar/Satellite
to Compose QPFNAM Forecast Boundary Layer
Moisture Convergence 18z June14 2005 (none over
OH River)
12
OH Valley Case Study-Using Models/Radar/Satellite
to Compose QPF 1719z Radar June 14 2005
13
OH Valley Case Study-Using Models/Radar/Satellite
to Compose QPF 1724z Satellite June 14 2005
14
Real Time Case Study-Short term QPFSatellite
Derived Convective Available Potential Energy-
June 14 2005 16z
15
Real Time Case Study-Short term QPFSatellite
Derived Lifted Index June 14 2005 16z
16
Real Time Case Study-Short term QPFSatellite
Derived Convective Inhibition June 14 2005 16z
17
Real Time Case Study-Short term QPFSatellite
Derived Precipitable Water June 14 2005 16z
18
OH Valley Case Study-Short term QPFJune 14 2005
Storm Total Precipitation
19
OH Valley Case Study-Short term QPFObserved 06
hour amounts ending 00z June 15 2005
20
Case Study Results

• NAM model diagnostics supported developing
  convection, but did not identify boundary to
  provide lift
• Satellite derived products supported model
  prognostics favorable for convection plus
  (combined with radar) identified boundaries to
  provide lift

21
Verification-How much Improvement Can We Derive
from Satellite/Radar/Model diagnostics?
22
Verification-24 Hour QPF vs. Models
23
FY2005 Verification
24
Short Term QPF Benefits from Multi-sensor Analysis

• Improved real time multi-sensor analysis would
• Reduce uncertainty of real time satellite/radar
  estimates
• Reduce uncertainty of post-event rainfall and
  time spent on quality control (more reliable
  verification)
• Lead to improvements in moisture/lift/instability-
  related diagnostics/prognostics, and thus
  confidence in qpf and excessive rainfall
  forecasts
• Questions/needed clarifications?

25
(No Transcript)
26
Composing a QPF

• Must have knowledge of
• Climatology
• Precipitation producing processes
• Sources of lift (boundaries, topography too)
• Forecasting Motion (propagation component vs.
  advection)
• Identifying areas of moisture/lift/instability

27
Analysis (Synoptic/Mesoscale)

• Perform a synoptic mesoscale analysis
• Upper air
• Upper fronts, cold pools, jet streaks
• Surface Data
• Boundaries
• Satellite Data
• Moisture plumes, Upper jet streaks
• Radar
• Boundaries
• Try to link ongoing precipitation with
  diagnostics

28
Analysis (Moisture)

• Precipitable Water (PW)
• Surface through 700 mb dew points
• Mean layer RH
• K indices
• Loops of WV imagery/derived PWs
• Consider changes in moisture
• Upslope/Down slope
• Veritical/Horizontal advection
• Soil moisture
• Nearby large bodies of water

29
Analysis (Lift)

• Low/Mid level convergence
• Lows, fronts, troughs
• Synoptic scale lift
• Isentropic
• QG components (differential PVA WAA)
• Jet dynamics
• Nose of LLJ
• Left front/right rear quadrants of relatively
  straight upper jets with good along stream
  variation of speed
• Mesoscale boundaries
• Outflow, terrain, sea breeze
• Orographic lift
• Solar heating

30
Analysis (Instability)

• Soundings are your best tool
• CAPE/CIN is better than any single index
• Beware!! Models forecast CAPE/CIN poorly
• Equilibrium Level
• Convective Instability
• Mid-level drying over low-level moisture
• Increasing low-level moisture under mid-level
  dry air
• Changing Instability
• Try to anticipate change from
• Low level heating
• Horizontal/Vertical temperature/moisture
  advection
• Vertical Motion

31
Precipitation Efficiency Factors

• Highest efficiency in deep warm layer
• Rainfall intensity is greater if depth of warm
  layer from LCL to 0o isotherm is 3-4 km
• Low cloud base
• Collision-Coalescence processes are enhanced by
  increased residence time in cloud
• Need a broad spectrum of cloud droplet sizes
• present from long trajectories over oceans
• Highest efficiency in weak to moderately sheared
  environments
• Some inflowing water vapor passes through without
  condensing
• Of the water vapor that does condense
• Some evaporates
• Some falls as precipitation
• Some is carried (blown) downstream as clouds or
  precipitation
• In deep convection, most of the water vapor
  input condenses

32
Low Level Jet

• Nocturnal maximum in the plains
• Inertial oscillation enhances the jet
• Often develops in response to lee low
  development
• LLJ can be enhanced by upper level jet streak
• Barrier jets (near mountains) can play a role in
  focusing lift
• Convection can induce very focused LLJs

33
LLJ Importance

• Speed convergence max at nose of LLJ
• Confluent flow along axis of the LLJ
• Vertical/Horizontal Moisture Flux positively
  related to strength of LLJ
• Differential moisture/temperature advection can
  lead to rapid destabilization
• Quasi-Stationary LLJ can lead to cell
  regeneration/training
• Often located on the SW flank of a backward
  propagating MCS

34
Movement of a system is dependent on cell
movement and propagation

The vector describing the propagation is the
vector anti-parallel to the LLJ Vprop
-VLLJ The vector that describes the movement of
the most active part of an MCS is represented by
V Vcell Vprop Propagation is dependent on
how fast new cells form along some flank of the
system
35
Factors leading to training/regenerating
convection

• Slow moving low level boundary
• Quasi-stationary low level jet
• Quasi-stationary area of upper level divergence
• Low level boundary (moisture/convergence) nearly
  parallel to the mean flow
• Lack of strong vertical wind shear (speed
  directional)