1999 Projects:
Information management and archiving
Data-sampling analysis
Project lead/manager: Daniel
Davis
Project team: Catherine Goyet
MBARI has developed several sophisticated systems capable of acquiring certain types of
data at relatively high rates in space and time, and the further development of these
systems and additional ones is likely. However, the inability to acquire other critical
measurements at a comparable rate in space or time can limit the scientific usefulness of
data available at increased rates. There can also be large mismatches between feasible
acquisition rates and the rates appropriate for answering important scientific questions.
Thus the successful development and use of such complex systems requires careful analysis
of the sampling strategy most appropriate to the scientific objectives.
The goal of this project is two-fold: to develop computational tools for analyzing
multidimensional data to determine sampling requirements for meeting scientific demands;
and, to develop appropriate sampling strategies when and where these requirements can be
met. The methodology to be applied is based on a technical approach that differs from
classical Shannon sampling theory, which was developed for one-dimensional time-series
data. Rather it is primarily focused on problems posed by multidimensional data acquired
at very low rates in space and time (such as bottle-sample data).
Next: Multi-resolution sampling and
estimation of oceanographic fields
Last updated: 07 October 2004 |
