Major physical processes that influence the forecast accuracy for the very short term

Jimy Dudhia, NCAR

The use of NWP models at very short range (0-6 hours) is considered to have the potential to improve
over extrapolative techniques in highly nonlinear situations. The most obvious of these is the development
of convective systems, but there may also be benefit in the interaction of weather systems with complex
terrain where models may be able to resolve high-resolution flow features fairly quickly after initialization.
For convective systems, the challenge is not just in the data assimilation to initialize a correctly
structured storm, but also in the physics to maintain a realistic development. Models with grid sizes
near 1 km can represent realistic severe convective systems well, but physics parameterizations are
known to affect their evolution. A particular example is in the microphysics where uncertainty in
particle size distributions lead to uncertainty in the evaporation rate, which in turn affects the
development or sustenance of the cold pool which impacts the further evolution and propagation of the
system within the first hours. The talk will address this and other physics uncertainties in the context
of short-range prediction.