Recommendations to the WorkshopBob Gall

• This workshop is highly focused – Could easily lead to the development of a

specific project• Objectives• Deliverables

– We should quickly settle on a structure and funding strategies

• Form a committee to oversee and guide the development of the project.

Constructively Competing Pressures?

• Real-time vs. retrospective• NDFD verification vs. broader applications• 2-D vs. 3-D• Deterministic vs. probabilistic• Microclimates vs. synoptic-mesoscale• Model independence vs. model dependence• now (Jan ‘05) vs. what can be done in a few

years• 1 analysis approach vs. several analysis

approaches

Classes of Analyses?• Provisional real-time analysis (PRTA):

– An analysis intended to provide within 30m-1h of the valid time the best possible specification of the state of the atmosphere subject to the limitations (data availability, retrieval, QC, etc.) inherent in creating an analysis within such a short time interval

– Potentially provides F00 NDFD products– PRTAs could be surface fields only or fully 3D

• Final real-time analysis (FRTA):– An analysis intended to provide the best possible specification of

the state of the atmosphere with particular attention placed on the conditions near the surface after at least a 1 day delay in order to allow for data latencies and additional QC

• Retrospective Analyses or Analysis of Record (AOR):– An analysis that uses the best available science to define the

state of the atmosphere

Is Dec ’04 Realistic for PRTA?

• Imposes constraints on process to build business model for AOR

• Eliminates benefits of potential collaborations with push to deliver single product

• Changing the NDFD verification paradigm for forecasters at the transition point

• At a minimum, data assets must be archived beginning December 1 to generate consistent retrospective analyses from the outset of the operational distribution of NDFD products

• What is realistic for EMC or FSL?• What resources are required?

The Long Term Target… Holy Grail

• AOR: An archive-quality (state-of-the-science) 3-D analysis of base-state and sensible weather elements. Multiple analysis strategies should be utilized as necessary to optimize quality of specific elements. Raw data assets will be archived centrally as a component of the AOR.

– Frozen through periods of parallel development– Documented error (uncertainty) characteristics– “system” transferred to operational cycle as appropriate

Stakeholders for AOR (courtesy F. Carr)

• Media –Broadcast –Print–Radio

• Energy–Power Generation, Transmission

and Distribution–Power traders

• Weather Derivatives Traders• Financial Institutions• Education• Research labs• Consumers• Recreational

• Traditional weather (& climate) community• North American Carbon Program • Air Quality (State EPA)• Homeland Security • Agriculture• Insurance • Urban Management• Transportation

– Aviation (Airlines, Corporate

Flight Depts, FBOs, Pilots, FAA)– Surface (Trucking, State DOT’s)– Marine (Shipping, Boaters)– Railway

IMPLEMENTATION WORKING GROUP

Expanding the Community Support for an AOR

• NDFD verification is a catalyst for mesoscale analyses, not an end in itself

• Develop compelling business case for an AOR

• Broaden awareness of the utility of an AOR

• Broaden funding base

Intensive Reanalysis of Extreme Seasons

A collaborative WRF and Climate community research project?

Craig BishopNaval Research Laboratory

Motivation for collaboration with Regional Climate Modeling

community• An increasing fraction of research funds is

being spent on climate research. Regional Climate Models (RCMs) need to accurately represent weather with high socioeconomic impact. Advanced high resolution mesoscale DA provides a powerful tool for diagnosing inaccuracies in the representation of such weather. Thus, if funds for climate research could be directed toward collaboration between RCM and mesoscale DA research both communities would benefit.

Activities

• Have regional climate researchers and mesoscale DA researchers collaboratively test their systems on extremes of historical weather.

• Climatologists identify extreme seasons of interest that are retrospectively analysed by WRF DA researchers.

• Climatologists and WRF DA researchers use advanced DA to diagnose model error at climate extremes.

Outcomes

• Better RCMs

• Better mesoscale DA

• RCMs based on WRF

• A WRF wrapper around state of the art mesoscale DA schemes

• Mesoscale reanalysis data for key periods

• Canned seasons with which to develop better DA, better WRF and better RCMs

First Steps

• Distribute survey of element-specific analysis capabilities. Responses needed ASAP.

• Post presentations in next couple of weeks to Web page (including ones from nonattendees, e.g., Chris Daly and Craig Bishop, or others who attended but were not able to present)

• Deliver executive summary of meeting to Jack Hayes/USWRP/meeting participants~ August 1

• Write a summary paper• Draft available to the community for comment by October 15• Final version submitted to BAMS ~ January 1

• Form a Mesoscale Analysis Committee

Mesoscale Analysis Committee • Preliminary charge:

– foster planning and implementation of high spatial and temporal resolution analyses of the atmosphere with particular attention placed on weather and climate conditions near the surface.

– Efforts should emphasize both real-time operational requirements and state of the science strategies to define an analysis of record.

– Coordinate with existing working groups within and external to NOAA/NWS

• Co-chairs? Brad Colman and John Horel• Order 10 members• Two subcommittees:

– Real-time analysis– Analysis of Record

• Composition should reflect community constituencies• Solicit volunteers and nominations. Deadline July 15• Recommendations submitted to Jack Hayes/USWRP by August 1

MAC Meetings

• First meeting: mid September– Recommend continued development of variational

and enKF for WRF– Recommend expandable sensor infrastructure and

additional sensors (e.g., pressure, wind, RH) be installed as part of COOP Modernization

– Assess and make recommendations regarding realistic options for PRTA

– Outline program plan for longer term R&D efforts and development of AOR

• Second meeting: mid April– Assess progress on PRTA– Finalize implementation plan for AOR

Realistic Timeline?• July 15: All presentations and preliminary recommendations available online.

Deadline for submission of names of volunteers and draftees for AOR committee membership

• August 1: Recommendations for committee members and executive summary of preliminary recommendations made to USWRP/Jack Hayes NWS

• ~Sept. 15: Committee meeting– Assess and make recommendations regarding realistic options for fast

track AORs– Outline program plan for longer term R&D efforts

• October 15: Draft paper/recommendations submitted for comment to community

• January 1: Final version of paper/ recommendations submitted for publication and to USWRP

• January 1: Preliminary examples of fast track AORs available• ~March 15: Committee meeting

– Assess progress on fast-track options– Finalize program plan for AOR R&D

• April 15: Rudimentary fast track AORs products available

Schedule• 8:30-8:45 Plenary session to review progress and goals

– What information is needed from other groups?• 8:45-10:15 Small group discussion wrap-up• 10:15-10:45 Break • 10:45-12:00 Summary of breakout sessions

– Each group will provide a few slides and summarize major issues & recommendations in 10 minutes with 10 minutes for discussion

• 12:00-1:00 Lunch (Cafeteria closed)• 1:00-3:00 Plenary discussion of AOR issues

– Strategy for implementation of recommendations – Recommendations on composition/organization of committee– The next steps: timeline

• 3:00 Adjourn

A Community Meeting on Real-time and Retrospective Mesoscale Objective Analysis:

An Analysis of Record Summit Brad Colman

• Can research and operations work together on this problem?

• Are there clearly definable requirements and objectives?

• Can we make a compelling argument?• If so, to look beyond here – what next?

Concerns

– Too much to handle– Not all constituencies represented at the

Workshop– AOR must have integrity – Need to avoid hasty implementation of a

deficient AoR scheme – Need to integrate WFO forecasters’

experience

Multiple Threads

RequirementsObservations

Approach

Jeff Anderson

Science, Technology, and ResourcesJohn Horel

• To what extent can the needs and requirements for objective analyses be met given existing scientific understanding, technologies, and resources?

• What are the critical scientific issues that must be faced in order to successfully develop quality analyses at high spatial/temporal resolution?

Breakout Session Details• Session 1: Analysis Requirements (Rm 3C404)

– Facilitators: David Ruth & John Snook– Charge: Identify and prioritize the requirements for AORs

• Session 2: Observational Resources and Verification (Room 3B809)– Facilitators: Jennifer Mahoney & Eugene Petrescu– Charge: Identify and prioritize the observational resources

required to generate AORs – Identify methods to assess the quality (verification) of the AORs

• Session 3: Analysis Strategies (Room GB124)– Facilitators: Fred Carr & Steven Lazarus– Identify and evaluate the current capabilities to develop AORs– Recommend ways that current deficiencies may be overcome

Breakout Session Suggestions

• Many potential issues to discuss in each session• Avoid tangential discussions and presentations• Consider issues broadly:

– What can be done now?– What do we need to know/improve for the future?– Real-time vs. restrospective issues– Local vs. national needs– Role of test beds

• Feel free to move between sessions– Brief break 4:10-4:15– Facilitators recap at 4:15

• If need subject matter expert, ask Brad or John to track one down

Meeting Outcomes?

• Formation of AOR Committee under NWS/USWRP auspices

• Fast-track AOR options

• R&D for the long term

Breakout Session 1: Analysis Requirements

• Charge: – Identify and prioritize the requirements for AORs

• Specific issues:– Can one analysis of record meet all needs?

• Local analyses vs. national products• Real-time vs. retrospective analyses

– What spatial and temporal resolution?– Averages vs. extremes in time/space? – Parameters (temperature, precipitation, etc.) – which

are absolutely essential now? – What are the fundamental characteristics that should

be met in the future?

Breakout Session 2: Observational Resources and Verification

• Charge: – Identify and prioritize the observational resources required to generate

AORs – Identify methods to assess the quality (verification) of the AORs

• Specific issues:– What observational data sets are most critical?– What limitations are imposed by the existing and future observational

data assets vs. those available in the past? – What is required to estimate quantitatively the observational errors in

the various data streams?– What efforts are required to collect surface data including mesonets?– How can COOP and other climate data be used effectively?– What is required to use radar and satellite data effectively?– How can the quality of analyses be assessed? – What are appropriate measures of analysis skill?– Can specific analytic experiments and case studies be defined to test

analysis solutions?

Breakout Session 3: Analysis Strategies

• Charge: – Identify and evaluate the current capabilities to develop AORs– Recommend ways that current deficiencies may be overcome

• Specific issues:– What are the strengths and weaknesses of existing methodologies?– What limitations are imposed by an underlying model? Sensitivity to

boundary layer parameterizations, soil moisture, clouds, etc.?– What are the realistic options for real-time and retrospective analyses

now?– Are there some aspects that are easier to accomplish than others?

Real-time vs. retrospective? Specific variables (temperature vs. precipitation)?

– What are the critical scientific issues that must be resolved in order to successfully develop quality analyses at high spatial/temporal resolution?

The First Step: Translating Needs to Requirements

• Many applications require the current and past states of the atmosphere near the surface at high spatial and temporal resolution.

• What does an analysis of record represent?• Can one analysis of record meet all needs?

– Local analyses vs. national products– Real-time vs. retrospective analyses– Resolution issues

• What spatial and temporal resolution?• Averages vs. extremes in time/space?

– Parameter issues (temperature, precipitation, etc.)

The Second Step: Science, Technology,

and Resource Inventory • To what extent can these requirements be met given existing

scientific understanding, technologies, and resources?– What can be learned from the literature and current applications?– What are the strengths and weaknesses of existing methodologies?– What observational data sets are most critical?– What limitations are imposed by the existing and future observational

data assets vs. those available in the past? – What limitations are imposed by an underlying model? Sensitivity to

boundary layer parameterizations, soil moisture, clouds, etc.?– What are the realistic options for real-time and retrospective analyses

within the next year or two?– Are there some aspects that are easier to accomplish than others?

Real-time vs. retrospective? Specific variables (temperature vs. precipitation)

– How can the quality of analyses be assessed? What are appropriate measures of skill?

The Third Step: Recommendations and Outcomes

• What are the critical scientific issues that must be faced in order to successfully develop quality analyses at high spatial/temporal resolution?

• What additional R&D and resources are needed?• Possible outcomes of this meeting?

– USWRP report based on recommendations from workshop and additional feedback from operational, research, and user communities

– Permanent NOAA/USWRP committee formed to continue advocacy for analysis of record

– Propose fast-track implementation plan to provide AOR soon– Develop long-term clearly-defined project plan with compelling

arguments – Propose funding opportunities and resources be sought to facilitate

research projects from NWS, NSF, and other agencies– Advocate long-term funding from NOAA/NWS and other agencies for

implementation of AOR beginning FY 07

USWRP Analysis of Record Summit Day 1: Tuesday, 29 June

• 8:00-8:15 AM Welcome (Bob Gall USWRP) • Session 1: Overview • 8:15-8:40 Analysis of Record Issues: Operational Perspective, Brad Colman, Seattle WFO. • 8:40-9:05 Analysis of Record Issues: Research Perspective, John Horel, University of Utah. • 9:05-9:25 Summary of Recommendations from USWRP Observing and Regional Weather Prediction Workshops.

Fred Carr, University of Oklahoma • 9:25-9:35 Summary of Recommendations from USWRP Cool Season QPF Workshop. Marty Ralph, NOAA ETL • 9:35-10:00 Break • Session 2: Current Capabilities • 10:00-10:15 MSAS/RSAS Patty Miller. FSL • 10:15-10:30

The NCAR/ATEC Real-Time Four-Dimensional Data Assimilation and Forecast (RTFDDA) System: Basics, operation and future development Yubao Liu. NCAR/RAP

• Session 3: Data Specific Applications • 10:30-10:45 Real-time Doppler wind quality control and analysis. Qin Xu NOAA/NSSL • 10:45-11:00 Challenges and Prospects for the Analysis of Precipitation Steve Vasiloff. NOAA/NSSL • 11:00-11:15 Real-Time Hurricane Wind Field Analysis Mark Powell. Hurricane Research Division • 11:15-11:25 Dave Sharp. Melbourne Fl WFO • 11:25-11:35 Robert Aune. NOAA/NESDIS • 11:35-12:45 Lunch • Session 4: Kalman Filter Applications • 12:45-1:00 Assimilation of Fixed Screen-Height Observations in a Parameterized PBL. Joshua Hacker NCAR • 1:00-1:15 Ensemble Filters for Data Assimilation: Flexible, Powerful, and Ready for Prime-Time? Jeff Anderson.

NCAR • 1:15-1:30 Toward a Real-time Mesoscale Ensemble Kalman Filter Greg Hakim. U. Washington • Session 5: Future Strategies • 1:30-2:00 Analysis of Record Strawman Proposal Geoff DiMego, NCEP. • 2:00-2:30 An FSL-RUC/RR proposal for the Analysis of Record Stan Benjamin. FSL • 2:30-3:00 A New Approach for Mesoscale Surface Analysis: The Space-Time Mesoscale Analysis System. John

McGinley. FSL • 3:00-3:15 Break