Research To Operations For Improved Precipitation Estimates and

Research To Operations For Improved Precipitation Estimates and Forecasts David Kitzmiller Group Leader - Hydrometeorology Hydrology Laboratory, Office of Hydrologic Development National Weather Service, NOAA Quantitative Precipitation Estimation Workshop June 28 -30, 2005 Norman, OK June 28, 2005 QPE Workshop - Norman

General Needs • Short-term precipitation estimates more than a few hundreds of meters from a rain gauge are generally poor • Better estimates are needed in order to realize improvements in: – – – June 28, 2005 Flash flood detection Distributed hydrologic modeling Verification Potential for wildland fires Agriculture QPE Workshop - Norman 2

Data Processing Needs • Problems with precipitation processing: – Erroneous rain gauge data – Artifacts in radar data – Variable precipitation detection by radar and satellite – Lack of knowledge of estimates’ error distributions June 28, 2005 QPE Workshop - Norman 3

“Inventors” for Improvements • Field office personnel – Devise and implement local solutions – Initiate requests for national support • Federal laboratory personnel – – NOAA NASA Dept. of the Interior Dept. of Defense • University community • Research foundations • Consultants June 28, 2005 QPE Workshop - Norman 4

Implementation Resources • Basic and applied research community amounts to 100’s of individuals worldwide • Number of implementers in NWS: – Office of Hydrologic Development: • 3. 5 for NEXRAD • 3 for AWIPS MPE – Office of Science and Technology: • < 10 in Systems Engineering Center, for NEXRAD and AWIPS • < 10 in Meteorological Development Laboratory, for AWIPS – Office of Operational Systems: • < 10 in Applications Branch, Radar Operations Center • Implementers in other NOAA organizations: – Forecast Systems Laboratory – NCEP Environmental Modeling Center – NCEP Central Operations Branch June 28, 2005 QPE Workshop - Norman 5

Implementation Resources • The people who implement new techniques must also modify and reengineer an ever-increasing catalog of existing applications • Staff for prioritizing new requirements and allocating resources is also small • Integrating staff must also prepare and sometimes deliver training materials, or innovations might go unused June 28, 2005 QPE Workshop - Norman 6

Basic Development – Any Developer – Part I • Develop new/improved products offline, using testbed server as data source • Get user buy-in – real-time demos, conference papers • Apply to Review Committee/User Grouip (to be formed) for incorporation in testbed • In testbed, demonstrate viability and utility of new product • Get further exposure: – NOAA and NWS personnel – General user community June 28, 2005 QPE Workshop - Norman 7

Basic Development – Any Developer – Part II • Review Committee reports routinely to NWS on new developments • An NWS sponsor will submit improvement proposal to OSIP – Proposal might not require remainder of OSIP process • ISSUES: – What are needs for JADE financial support as the main testbed? – Getting NWS and community support for JADE – How do we form/maintain Review Committee or advisory group? June 28, 2005 QPE Workshop - Norman 8

NMQ Activities At NCEP • Basic 3 -D mosaic process is under testing at NCEP EMC • Following must go through OSIP: – Justify 3 -D mosaic generation and distribution – Justify need for production of Q 2 (precipitation) for: • RFC needs • WFO needs • AOR needs – Concept of Operations for Q 2 – Must document continuing resources needs: • Supporting internal coding support • User support • 24/7 troubleshooting support June 28, 2005 QPE Workshop - Norman 9

NMQ Implementation Issues • Who starts as OSIP sponsor to write Statement of Need, Concept of Operations? • Exactly where to implement? – Centrally? – Regionally? • Who provides short-term science support? – Network Control Facility? • Who provides long-term science support? • Benchmarks for approval for implementation? – Retrospective verification over an annual cycle? – Stage II and/or Stage IV as benchmark? – Final AOR as longterm benchmark? June 28, 2005 QPE Workshop - Norman 10

Validation of Improvements • Retrospective verification over an annual cycle • Regional as well as national validity • Get benchmark for existing Stage II product as a benchmark • How to bring in RFC quality control and expertise operationally • Final AOR as longterm benchmark June 28, 2005 QPE Workshop - Norman 11