3rd NSTWS, Vienna VA, July 25-27 2007 Research to Operations in the Joint Center for Satellite Data Assimilation Lars Peter Riishojgaard Director, JCSDA.

3rd NSTWS, Vienna VA, July 25-27 2007

Research to Operations in theJoint Center for Satellite Data

Assimilation

Lars Peter Riishojgaard

Director, JCSDA

3rd NSTWS, Vienna VA, July 25-27 2007

Overview

• JCSDA; mission and partners

• Research to Operations

• Failure modes

• What next ?

3rd NSTWS, Vienna VA, July 25-27 2007

What is the JCSDA?

• An interagency organization set up by NASA (GSFC/GMAO), NOAA (NCEP/EMC, NESDIS/STAR, OAR/ESRL), and DoD (AFWA, NRL/Monterey)

• JCSDA Mission: … to accelerate and improve the quantitative use

of research and operational satellite data in weather, ocean, climate and environmental analysis and prediction models

3rd NSTWS, Vienna VA, July 25-27 2007

What does the Joint Center do?

• Primarily Numerical Weather Prediction on the global scale

– Numerical models run out to 7-10 days– Horizontal resolution ~50 km– Vertical resolution (in the atmospheric boundary

layer): hundreds of meters

3rd NSTWS, Vienna VA, July 25-27 2007

Why should you care about global NWP?

• Provides boundary conditions to regional/mesoscale models used for surface weather

• Data assimilation is a great technique for “data homogenization”

• Longer range forecasts require global approach

3rd NSTWS, Vienna VA, July 25-27 2007

NWP observational data requirement as a function of forecast range

Improvement in forecast skill formiddle and high latitudes

Polar winds from MODIS:Pioneered by GMAO, transferred via JCSDA to NCEP/EMC

• Sequences of water vapor images

from MODIS can be used for wind retrivals in high-latitude regions not seen from geostationary orbit

MODIS WV imagery, 24 h periodConsistent 15-20% improvement inhurricane track forecasts in 2004

3rd NSTWS, Vienna VA, July 25-27 2007

Research to Operations; why do we (occasionally) fail?

I. Lack of funding

II. Lack of coordination between R and O

III. Lack of human resources

IV. Premature exit

V. Science not sufficiently mature

VI. Lack of awareness of needs

3rd NSTWS, Vienna VA, July 25-27 2007

Scientific expectation gap: the case of

Quikscat

3rd NSTWS, Vienna VA, July 25-27 2007

Scientific roadblocks

• Imagery data/feature tracking winds– Quality control and height assignment notoriously

difficult

• Surface winds– Not a model variable; difficult to relate to

prognostic variables

• Anything related to the hydrological cycle, e.g. clouds, precip– Arguable the most important unsolved scientific

problem in atmospheric science

3rd NSTWS, Vienna VA, July 25-27 2007

What can we do here?

• Open the dialog!

• JCSDA needs to be aware of needs of all users, direct as well as indirect - including Surface Transportation communities

• JCSDA user communities need to be cognizant of NWP capabilities and plans, and have a voice in the formulation of the latter