Transcript DTC Task Lead Meeting - July 2010 - Boulder, CO

Evaluation of QPF during the HMT-West Winter
Exercise: A DTC/HMT Collaboration with USWRP
Edward Tollerud1, Tara Jensen2, John Halley Gotway2, HuilingYuan1,3, Wally
Clark4, Ellen Sukovich4, Paul Oldenburg2, Randy Bullock2, Gary Wick4 (also
Tressa Fowler, Barb Brown, Matt Pocernich, Paul Schultz, Chris Harrop, Isidora
Jankov)
1
ESRL/GSD, Boulder, CO
3
CIRES, Boulder, CO
http://esrl.noaa.gov/gsd/fab
2
4
NCAR/RAL, Boulder, CO
ESRL/PSD, Boulder, CO
Wally Clark
USWRP/Test-bed Workshop 2010
HMT-DTC/USWRP Collaboration
Collaborating Agencies and Motivations
GSD: High-resolution ensemble NWP and verification
PSD, USWRP: Atmospheric Rivers and QPF; Severe events
NCAR/RAL: MET-based real-time and object–based verification; HWT
interactions
RFCs, WFOs, CA DWR: Operational reservoir management
Goals and Objectives
 Implementation and demonstration of new verification
capabilities for high-resolution NWP
 Develop DTC capabilities for ensemble model probabilistic
prediction and verification
 Data Impact Studies
 Impact studies of model physics and parameterizations
relevant to HMT research
Duality of Purpose: A Tale of Two Communities
1) Research for the USWRP, HMT, and etc.
Very strong initial focus on QPF and extreme precipitation
Important observational component
What physical mechanisms lead to damaging rainfall?
What verification tools can provide valuable diagnostic information?
(eg. MODE vis a vis atmospheric rivers)
2) Operational Forecast Guidance for RFCs, CDWR, NWS, and etc.
Baseline comparison with previous products (i.e, GFS forecasts)
Establish usefulness of high-resolution WRF ensemble forecasts
What is value of ensemble mean?
What are limits to forecast lead times for QPF?
HMT Winter Exercise – Online
Real-time Verification
1) MODE and standard scores
2) Selection of Vx data: gages, Stage
IV, daily, 6h
3) Choice of format: valid time,
initialization time, monthly
summary, time series, bar plots,
boxplots, variable thresholds,
regional selections
USWRP/Test-bed Workshop 2010
Real-time QPF verification for HMT-West
• Initialized 12 UTC 1/17/2010, Verif. Stage IV, 0.1 in. threshold
• Good performance for ensemble mean (black ) and GFS (brown)
• Better performance overall for heavy rainfall periods
• http://verif.rap.ucar.edu/eval/hmt/2010/graphics/
18 January 2010
19 January 2010
Wind turning: diurnal forcing or pure
happenstance??
USWRP/Test-bed Workshop 2010
Real-time QPF verification for HMT-West
• Constant Valid Time:
1200 UTC 01/20/2010
• Spatial Frequency Bias,
.01 inch threshold
• Full domain, .01 threshold
• Verified with 6h Stage IV
•Not large change with lead
time except longest (4-5
day)
•Anomalously Large bias
for ensemble mean
USWRP/Test-bed Workshop 2010
Real-time QPF verification for HMT-West
• January Summary, ETS
•24h forecast, full domain
•Verified at 24h gage sites
• Segregated by threshold
• Dramatic performance hit
at higher thresholds
•No real ensemble member
advantages in statistics but
large intger-quartile range
USWRP/Test-bed Workshop 2010
Real-time QPF verification for HMT-West
• January Summary, FAR
•24h forecast, full domain
•Verified at gage sites
• Segregated by threshold
•Good Performance by
Ensemble Mean, poor for
GFS
PODY verified at points with
24h gage observations
PODY verified at Stage
IV grid points (6h)
Ensemble Members, Ensemble Mean, and Observed (Stage IV)
Precipitation Fields – 6h forecasts valid at 1200 UTC 20 January 2010
Ensemble Members, Ensemble Mean, and Observed (Stage IV)
Precipitation Objects – 6h forecasts valid at 1200 UTC 20 January
2010
Object-based Atmospheric River Verification of Integrated Water Vapor
72h
Area=369
http://esrl.noaa.gov/gsd/fab
MODE Object Comparison of
48h
24h
GFS Forecasts
with SSM/I Observation
for 25 February, 2004 (Clear Cut Case)
Area=312
Area=306
SSM/I
Area=127
Wally Clark
USWRP/Test-bed Workshop 2010
Proposed Future Activities
 Expand and maintain real-time QPF verification demonstration system
and analyze results (HMT-West and eventually HMT-East)
 Develop and assess prototype MODE-based spatial verification
techniques to verify Eastern Pacific Atmospheric River events in nearreal-time
 Addition of new ensemble-based probabilistic techniques to the MET,
including spatial ensemble verification prototypes
 Investigate data impacts WRT verification data sets and initialization
data
Prioritized Tasks: Details
 Adapt HWT-based probability scoring for HMT
 Incorporate GFS into skill scores for base-lining
 Add SREF to ensemble set
 Display Quantitative attributes from MODE
 Add wind verification for HMT-West
 Build 6h gage verification option
 Develop MODE-based ensemble verification utilities
 Include Variable duration option for verification periods
Prioritized Tasks: Details (Continued)
 Develop Credible Downscaling Applications for model Comparisons
 Extend Metviewer to Other Menu-Driven Procedures
 HMT-East Prototype Products
 Introduce Time-series Verification Options
 Publication of Winter Exercise Studies
Collaboration/Leverage with other DTC Tasks and
Testbeds
 Probability and the HWT
 Verification of winds: QNSE
 Metviewer and Ensemble Processing for MET
 Ensemble Product Development, post-processing with DET
 HMT, ARs, and MET/MODE
Presentations and Publications
 Meeting/Workshop Presentations:
 BACIMO Workshop (Halley Gottway et al.)
 WRF User’s Conf. (Tollerud et al.; Halley Gottway et al.)
 TestbedWorkshop (Tollerud et al.; Clark et al.)
 HMT West Tutorial (Jensen and Tollerud)
 Verification Colloquium (Tollerud et al.)
 Presentations planned for Future
 Weather Analysis and Forecasting; Hydrology
 QPF3 and GRAPES
Formal Publication Plans
• Diagnostic Results from a Real-time QPF Verification System for the
HMT Winter Exercise
• Assessing the Performance of a High-resolution WRF ensemble
Forecast System Relative to GFS Forecasts
• Object-based Verification for Historical Pacific Atmospheric Rivers
• Techniques for Spatial Verification using IWV and moisture Flux
Objects
• QPF Verification Uncertainty and Other Impacts of Precipitation
Dataset Choice