JEFS Project Update

1
JEFS Project Update And its Implications for the
UW MURI Effort
Cliff Mass Atmospheric Sciences University of
Washington
2
ENSEMBLES AHEAD
JEFS
3
Joint Ensemble Forecast System(JEFS)
NCAR
4
JEFS Goal
Prove the value, utility, and operational
feasibility of ensemble forecasting to DoD
operations.
5
J E F S T E A M
AFIT
6
Joint Global Ensemble (JGE)

• Description Combination of current GFS and
  NOGAPS global, medium-range
• ensemble data. Possible
  expansion to include ensembles from CMC,
• UKMET, JMA, etc.
• Initial Conditions Breeding of Growing Modes 1
• Model Variations/Perturbations Two unique
  models, but no model perturbations
• Model Window Global
• Grid Spacing 1.0?? 1.0? (80 km)
• Number of Members 40 at 00Z
• 30 at 12Z
• Forecast Length/Interval 10 days/12 hours
• Timing
• Cycle Times 00Z and 12Z

1 Toth, Zoltan, and Eugenia Kalnay, 1997
Ensemble Forecasting at NCEP and the Breeding
Method. Monthly Weather Review Vol. 125, No.
12, pp. 32973319.
7
Joint Mesoscale Ensemble (JME)

• Description Multiple high resolution,
  mesoscale model runs generated at FNMOC
• and AFWA
• Initial Conditions Ensemble Transform Filter 2
  run on short-range (6-h),
• mesoscale data
  assimilation cycle driven by GFS and NOGAPS
• ensemble
  members
• Model variations/perturbations
• Multimodel WRF-ARW, COAMPS
• Varied-model various configurations of physics
  packages
• Perturbed-model randomly perturbed sfc
  boundary conditions (e.g., SST)
• Model Window East Asia
• Grid Spacing 15 km for baseline JME
• 5 km nest later in project
• Number of Members 30 (15 run at each DC site)
• Forecast Length/Interval 60 hours/3 hours

7 h production /cycle
2 Wang, Xuguang, and Craig H. Bishop, 2003 A
Comparison of Breeding and Ensemble Transform
Kalman Filter Ensemble Forecast Schemes. Journal
of the Atmospheric Sciences Vol. 60, No. 9, pp.
11401158.
8
UW MURI Contributions

• UW team making major contributions to the JEFS
  mesoscale system including
• Observation-based bias correction on a grid
• Localized BMA
• Work on a variety of output products

9
NCAR Contributions

• Ensemble Model Perturbations
• a. Improvement of multi-model approach (0.5 FTE)
• The current method to account for model
  uncertainty in the JME, developed by NCAR in
  FY06, includes a multi-model component (i.e.,
  each ensemble member represents a unique model
  configuration or combination of physics schemes)
  and perturbations to the surface boundary
  conditions (SST, albedo, roughness length,
  moisture availability). This method will be
  further improved by the following additions.
• 1) Incorporation of additional physics schemes.
• 2) Tuning of sea surface temperature (SST)
  perturbation.
• 3) Addition of soil condition perturbation. (0.25
  FTE)

10
NCAR Contributions

• Development of new approaches
• 1) Multiple-parameter (single-model) approach.
• NCAR shall examine the representation of model
  uncertainty through the use of a single, fixed
  set of model physics schemes in which various
  internal parameters and "constants" of each
  scheme are varied among the ensemble members.
• 2) Stochastic-model approach.
• NCAR shall adapt to WRF a stochastic modeling
  approach (stochastic physics or stochastic
  kinetic energy backscatter).
• 3) Hybrid approach. As the most straightforward
  hybrid method, NCAR shall apply the developed
  stochastic-model approach on top of the
  multi-model approach.

11
NCAR

• Evaluation of approaches (0.4 FTE)
• MMM shall evaluate the different approaches for
  diversity that properly represent model
  uncertainty.
• Determination of best approach and assistance
  with implementation

12
UW Contributions 2007

• Ensemble Post-processing Calibration
• The University of Washington Atmospheric
  Sciences Department (UW) on developing algorithms
  for post-processing calibration of mesoscale
  ensembles. This development effort is crucial
  for optimizing the skill of ensemble products and
  maximizing JME utility. The UW shall
• a. Expand model bias correction. The
  observation-based, grid bias correction developed
  in FY06 for 2-m temperature will be extended to
  additional variables of interest to include, but
  not be limited to, 2-m humidity, 10-m winds, and
  cumulative precipitation (rain and snow).
• b. Develop ensemble spread correction. The
  prototype Bayesian Model Averaging (BMA)
  post-processing system developed in FY06 shall be
  fully developed for the same variables as noted
  for bias correction.
• c. Evaluate developments. The UW shall evaluate
  these calibration techniques to determine the
  gain in ensemble forecast skill.

13
UW JEFS

• 3.3 Ensemble Products and Applications
• For FY07, NCAR/MMM shall continue subcontract
  work with UW on developing JME products and
  applications. The UW, under direction of NCAR,
  shall develop the following prototypes. These
  deliverables are initial efforts that do not
  require delivery of finalized software and
  documentation.
• a. Extreme forecast index. The UW shall research
  state-of-art methods for calculating an
  ensemble-based extreme forecast index and develop
  a prototype capability for the JME. This
  essentially is the process of comparing the
  current ensemble forecast with the ensemble
  models climatology to determine the likelihood
  of an extreme event, one that might not even be
  represented within the ensemble.
• b. General user interface. The UW shall build a
  web-based, interactive JME interface for the
  general DoD user designed to provide basic
  stochastic weather forecast information. This
  will be similar in nature to the current Probcast
  interface (http//www.probcast.com/) except
  geared to address the specific interests of
  military operations (e.g., probability of low
  ceiling and visibility).

14
UW Contributions

• The UW team will expand in 2007 to include
  several members of the UW Statistics Deparment.
• Potential for further expansion in FY 2008.

15
Product Strategy
Tailor products to customers needs
and weather sensitivities Forecaster
Products/Applications ? Design to help transition
from deterministic to stochastic
thinking Warfighter Products/Applications ?
Design to aid critical decision making
(Operational Risk Management)
UW will aid in developing some of these products
16
Operational Testing Evaluation
PACIFIC AIR FORCES Forecasters 20th
Operational Weather Squadron 17th
Operational Weather Squadron 607 Weather Squadron
Warfighters PACAF 5th Air Force
Naval Pacific Meteorological and Oceanographic
Center Forecasters Yokosuka Navy
Base Warfighters 7th Fleet
SEVENTH Fleet
FIFTH Air Force
17
Forecaster Products/Applications
18
Consensus Confidence Plot
Maximum Potential Error (mb, /-)
6 5 4 3 2 1 lt1

• Consensus (isopleths) shows best guess
  forecast (ensemble mean or median)
• Model Confidence (shaded)
• Increase Spread in
  Less Decreased confidence
• the multiple forecasts
  Predictability in forecast

19
Probability Plot

• Probability of occurrence of any weather
  phenomenon/threshold (i.e., sfc wnds gt 25 kt )
• Clearly shows where uncertainty can be exploited
  in decision making
• Can be tailored to critical sensitivities, or
  interactive (as in IGRADS on JAAWIN)

20
Multimeteogram
21
Sample JME Products
Probability of Warning Criteria at Osan AB
When is a warning required?
What is the potential risk to the mission?
Valid Time (Z)
Surface Wind Speed at Misawa AB
Extreme Max
Requires paradigm shift into stochastic thinking
Mean
90 CI
Extreme Min
11/18 12/00 06 12 18
13/00 06 12 18
14/00 06
Valid Time (Z)
22
Warfighter Products/Applications
23
Bridging the Gap
Stochastic Forecast
Binary Decisions/Actions
AR Route Clear 7
Go / No Go
T-Storm Within 5
?
IFR / VFR
GPS Scintillation
Bombs on Target
Crosswinds In / Out of Limits
Flight Hazards
24
Method 2Weather Risk Analysis and Portrayal
(WRAP)