8
NEVILLE SMITH
One premise is that GODAE is not just concerned with prediction in the
traditional sense (looking forward in time), but prediction in its most general
form, where information is extrapolated and interpolated in space and time,
and between fields (Figure 1). The objectives intentionally imply a broad
scope in the belief that wide utilization and exploitation of products are
essential for cost-efficiency and relevance to society.
Figure 1. Schematic of the processes used to exploit data. In some cases we use linear,
perhaps empirical relationships to relate the current state to, say, a likely future state. In other
cases forecasts are produced based on current data (“today”), perhaps at a specific location
(“here”), and perhaps for a subset of the total variable space (“ours”), in order to forecast the
state in the future (“tomorrow”), at some remote location (“there”) or for some variables that
are not part of the observables (“yours”). The process involves extrapolation (e.g., a forecast),
interpolation (e.g., discrete points to a grid) and interpretation (e.g., inferring winds from sea
surface topography).
The strategy for the development of these products is built on the concept
of a GODAE “Common” which is shared by all GODAE Partners
responsible for realizing the goals and objectives of GODAE. The GODAE
Common concept is essential for GODAE, and must also be transported into
the “operational” environment, for example through data policies and
scientific cooperation.
5.
Building the systems
The essential building blocks of GODAE are observations, models and
estimation tools (Figure 2). In the GODAE context, these elements are
inextricably linked, with obvious two-way interdependencies. The
generation of globally consistent fields of ocean temperature, salinity and
velocity components through the synthesis of multivariate satellite and in
situ data streams into ocean models is a central theme of GODAE.
NEVILLE SMITH
One premise is that GODAE is not just concerned with prediction in the
traditional sense (looking forward in time), but prediction in its most general
form, where information is extrapolated and interpolated in space and time,
and between fields (Figure 1). The objectives intentionally imply a broad
scope in the belief that wide utilization and exploitation of products are
essential for cost-efficiency and relevance to society.
Figure 1. Schematic of the processes used to exploit data. In some cases we use linear,
perhaps empirical relationships to relate the current state to, say, a likely future state. In other
cases forecasts are produced based on current data (“today”), perhaps at a specific location
(“here”), and perhaps for a subset of the total variable space (“ours”), in order to forecast the
state in the future (“tomorrow”), at some remote location (“there”) or for some variables that
are not part of the observables (“yours”). The process involves extrapolation (e.g., a forecast),
interpolation (e.g., discrete points to a grid) and interpretation (e.g., inferring winds from sea
surface topography).
The strategy for the development of these products is built on the concept
of a GODAE “Common” which is shared by all GODAE Partners
responsible for realizing the goals and objectives of GODAE. The GODAE
Common concept is essential for GODAE, and must also be transported into
the “operational” environment, for example through data policies and
scientific cooperation.
5.
Building the systems
The essential building blocks of GODAE are observations, models and
estimation tools (Figure 2). In the GODAE context, these elements are
inextricably linked, with obvious two-way interdependencies. The
generation of globally consistent fields of ocean temperature, salinity and
velocity components through the synthesis of multivariate satellite and in
situ data streams into ocean models is a central theme of GODAE.