PPAV Team 2014 Updatesand

Transitioning Research to Operations

David Zelinsky, Mark DeMaria, Tim Marchok

PPAV Team Leaders

PPAV Team

Jim Hansen (NRL)John Knaff (NESDIS/STAR)Gopal (AOML) T.N. (Krish) Krishnamurti (FSU) Jeff Hawkins (NRL) Thiago Quirino (AOML) James Franklin (NHC) Paul Kucera (NCAR/TCMT)Jack Beven (NHC) Kevin Yeh (AOML) Andrew Penny (NHC) Jun Zhang (UM CIMAS/AOML)

Mark DeMaria (NHC) David Zelinsky (NHC)Tim Marchok (GFDL) Vijay Tallapragada (EMC) Sim Aberson (AOML) Kate Musgrave (CIRA/CSU) Jon Moskaitis (NRL) Buck Sampson (NRL) Robert Rogers (AOML) Barb Brown (NCAR) Richard Pasch (NHC) Michael Fiorino (ESRL) Louisa Nance (NCAR)

Outline Brief updates on 2014 progress If you have questions about any of the work presented, please note the agency on

each item

Priorities for 2015 (and beyond) Comments on transitioning Research to Operations

Contributing Organizations NHC NRL EMC/HWRF DTC HRD NESDIS/CIRA GFDL TCMT

NHC Updates Continued investigating if global model forecasts can be used for intensity forecasting Continued support of the HFIP products webpage Maintained NHC’s internal operational model diagnostics intranet page Began running the DTC MET-TC verification tool at NHC. Real-time verification

graphics were provided to the forecasters Continued diagnostic support of the operational models, especially HWRF Implemented the 5 day Graphical Tropical Weather Outlook on the NHC website,

including gis data Began development of a new weighted track and intensity consensus model for

NOAA Hired new model diagnostician and GIS specialist Updated tropical cyclone watch/warning breakpoint maps and data on the NHC

website

NHC Updates

5 day Graphical Tropical Weather Outlook and associated GIS data

Initial development on HCCA, HFIP corrected consensus track and intensity

NRL Updates ATCF at NHC was upgraded to the latest version (5.6) A summary of improvements is available here: http://www.nrlmry.navy.mil/atcf_web/docs/requirements/changed_in_560.txt

HWRF updates Established the 2014 baseline HWRF and tested several major upgrades prior

to the 2014 hurricane season. Conducted regular model diagnostics on all tropical cyclone basins including

the W. Pac and Indian Ocean during the hurricane season and maintained the HWRF forecast webpage

Development of Rapid intensification verification techniques (more on DTC slides)

Performed conditional verification of track and intensity forecasts based on various criteria

Beginning evaluation of HWRF applications for Wave, Surge, and Hydrology models

Began in-house evaluation of tornadic potential fields from HWRF for landfalling storms.

DTC Updates

Evaluation of Rapid Intensification/weakening• Methodology has been developed in consultation with HFIP partners• The code to evaluate RI/RW has been added to MET-TC

• Computes Probability of Detection , False Alarm Rate etc.• Calculates RI/RW performance with user-defined intensity and time thresholds• Stratifies performance based on initial intensity, location and size of the storm• Allows for homogeneous comparison among several models• Relates RI errors to intensity errors

The code developed will be available to the community next year

DTC Updates Results: Evaluation of Rapid Intensification/Weakening H214 (2014 HWRF) POD is ~25% or lower in AL and EP, although it increases to

0.47/0.30 (AL/EP) if the forecast intensity change criteria is relaxed from 30 to 25 kt. H214 has a higher false alarm ratio in the AL than EP.

HRD Updates Continued development of new diagnostics for HWRF Vortex-scale Convective-scale PBL and surface layer

Conducted the Hurricane Field Program summer visitor program 2014 visitors: Hui Liu, Jonathan Vigh, Shuyi Chen, Annette Foerster, Heather Holbach

HRD Updates Comparisons of 122 HWRF forecasts of 4 hurricanes (Bill

2009, Earl 2010, Karl 2010, Irene 2011) with large Km (PBL11) and reduced Km more consistent with observations (PBL12)

Distribution of intensity for two composites

• Vortex-scale diagnostics:

• PBL12 (reduced Km) produces a larger proportion of stronger storms than PBL11

PBL diagnostics:• PBL12 produces a shallower

inflow layer than PBL11, more consistent with dropsondes

NESDIS/CIRA Updates Continued development of experimental SHIPS, LGEM, and SPICE Experimental ECMWF-based SHIPS and LGEM SPICE updated with 2014 coefficients Evaluation of DSHP/LGEM weighting in SPICE

Continue development of the large-scale (SHIPS) diagnostic code and associated products Realtime verification products are on the CIRA website

Performed verification of large-scale HWRF synthetic imagery and model moisture fields using archived GOES data and total precipitable water

Hybrid windspeed probability grids and graphics generated and displayed on the HFIP webpage

NESDIS/CIRA Updates Verification of large-scale total

precipitable water Underway for the 2014 HWRF

Comparison and difference plots and normalized histograms for each forecast.

Seasonal basin wide, numbered storm comparison generated. Biases and MAE per grid point.

GFDL Updates GFDL New GFDL tracker implemented operationally and released to the community. Wind radii GFDL tracker issue fixed Evaluations performed on ~20 different configurations of an upgraded GFDL model,

including new Cd and Ch formulations

GFDL Updates New GFDL tracker implemented operationally at NCEP with the 2014 HWRF

package and released to the community by DTC in August 2014 Wind radii GFDL tracker issue has been fixed

The biases in the forecast 34-kt wind radii values reported by the tracker in the NE quadrant for Hurricane Sandy are now closer to zero.

TCMT Updates Currently developing metrics to quantify the run-to-run consistency of

forecasts

NHC Display and Diagnostic System Developing an interactive Real-time display and diagnostic system for the display of

model forecasts (initially, tracks and intensity)

TCMT Updates

Track forecasts were consistently to the left of the observed track, but were close to the observed track at shorter lead times

Initial track forecasts were consistently to the right of the observed track, but were close to the observed track at shorter lead times

PPAV priorities for the next 5 years of HFIP

Development of new products & statistically post-processed guidance Development of new diagnostic & verification techniques Real-time display, diagnostics, and verification Community tool development

FY15 Priorities

Implement the NHC-HFIP Corrected Consensus in operations Develop techniques that provide guidance on guidance Upgrade the CIRA model diagnostic code Enter CIRA model diagnostic code into DTC as a community code Develop a set of standard metrics and verification techniques that assess surface wind structure and storm size. Continue HRD vortex- and convective-scale diagnostics (e.g., RMW, vortex tilt, eyewall slope, convective burst radial

and azimuthal locations, CFADs of vertical velocity) Develop air-sea interaction diagnostics (e.g., continue HRD PBL & surface layer diagnostics, surface fluxes, ocean SST,

MLD, OHC, currents, upwelling) Include new DTC/TCMT verification techniques into MET-TC (e.g., forecast consistency verification, RI, genesis

verification) Release latest version of GFDL vortex tracker to the community Evaluate HFIP website for content and revisions Create a link on the HFIP web page of the list of available verification and diagnostic tools from all the various groups. Develop a procedure for evaluating web-based HFIP products and transferring high-priority items to NHC Coordinate with NRL to develop a method for creation of an itemized list of NHC requirements for ATCF support

and upgrades, including a budget consistent with these tasks. Work with NRL to implement the changes.

Transition of HFIP Applications to Operations HWRF Forecast Improvements Direct path to NHC through ATCF

NHC Guidance Suite SHIPS/LGEM/RII improvements on WCOSS

ECMWF version of SHIPS/LGEM Wind speed probability model Transition of new statistical post-processing to WCOSS

SPC3 (K. Musgrave), HCCA (A. Simon) TC genesis forecasts to WCOSS or local NHC platforms

Model Diagnostics Transition to NHC operational platforms for HSU use

N-AWIPS, AWIPS, ATCF, NHC compute serves Advanced verification systems for HWRF evaluation

Improvements to public-facing products on NHC web page GIS product development

Guidelines to Facilitate Operational Transition-1 Separate I/O from processing Allows for data input modules and user interfaces/graphics to swapped with

operational versions

Use standard formats as much as possible grib, NetCDF, ATCF, BUFR, etc

Avoid software the requires licensing MATLAB, IDL, some graphics programs

Preferred Linux scripts, Python, PERL, FORTRAN, C, C++, Java

Guidelines to Facilitate Operational Transition-2 Automation essential Consider long-term maintenance Provide capability for algorithm updates

Coordination with NHC for promising algorithms Include a few in HFIP milestones?

NHC TSB staff is spread thin