RONALD M. ERRICO
The National Center for Atmospheric Research
1850 Table Mesa Drive
Boulder, Colorado 80307, USA
Abstract:
The basic principles of modern data atmospheric data assimilation
will be described. The intent is, in some sense, to optimally combine
diverse sets of observations and information (e.g., physical
relationships) accounting for the fact that all is uncertain to
various degrees. It is a fundamentally statistical problem, but
computationally intractable for most standard techniques due to the
size of the problem and the uncertainty even of the prior statistics.
Current techniques will be described along with their deficiencies.
Mathematical, statistical, and physical requirements for making
further progress will also be described. Some common misconceptions
will be revealed.