Progress in testing hypotheses for the North American Observing System (NAOS) Program

ed.The NAOS Program exists to develop cost-effective strategies for deploying atmospheric observing systems in support of weather and climate services. The Program has generated a number of hypotheses regarding present and future observing systems with the intent of providing a scientific rationale for decision-making. Results of hypothesis testing are now becoming available. For example, wintertime tests suggest that rawinsonde data at fourteen U.S. locations are redundant with nearby ascent/ descent soundings from commercial aircraft for the purpose of numerical forecasting. Targeted observations (dropsondes) in the eastern North Pacific Ocean have led to improved North American late winter forecasts in most cases, though counterexamples suggest that assimilation practices need improvement and the entire area of sensitivity at the initial time must be sampled. Targeted observations surrounding landfalling tropical storms have led to improved track forecasts. Despite these early successes, the basic challenge still facing NAOS is to conduct tests in advance of major decisions on observing systems and to link results of these tests to the decisions themselves. NASA has provided most of the computing resources until recently. Now NCEP and FSL have acquired additional computing resources dedicated to NAOS testing, and the pace of testing will accelerate. The schedule for testing future hypotheses will be describ