Spatial Temporal Events
Original Publication Date: 1993-Jun-01
Included in the Prior Art Database: 2005-Mar-19
An event is an object that has spatial and temporal extents and has meaning to a user. A scientist observing a data set is searching for criteria to formulate models, i.e., causal relationships between the data at different times. This requires the ability to define the events in the GIS database so they can be further analyzed.
Spatial Temporal Events
An event is
an object that has spatial and temporal extents and
has meaning to a user. A scientist observing a data set is searching
for criteria to formulate models, i.e., causal relationships between
the data at different times. This requires the ability to define the
events in the GIS database so they can be further analyzed.
Relationships among events are used to define conditions
combining events to form a hierarchy of "higher order" events.
The following are some examples of events:
o Growing Season
- Some floods in the Pacific Northwest are caused
by snow followed by rain, and it is hypothesized that this
process is exacerbated by forest clear cuts. Years of data are
available, and clear cuts, precipitation, runoff, and floods
could examined for causal relationships. A clear cut event could
have attributes associated with it, such as area, tree type
before the cut, and if reforestation plays a role, the degree of
canopy could also be a time series attribute of the clear cut. A
rain storm event could have water volume as an attribute. Upon
investigating the relationships between the clear cut, rain
storms and flood events, statistical as well as spatial
relationships could be taken into account.
o Cloud Cover - Cloud cover could be recognized by a drop followed
by a rise in green biomass detected from a remote sensor. This
would differentiate it from a harvest or locust swarm which would
involve a permanent drop in green biomass. The existence of a
cloud cover event could be corroborated by coincident and
concurrent storm events.
o El Nino - The study of El Nino seems to be particularly suited
for the event methodology. Already many diverse concurrent
events to the warm water off Peru are studied, such as Southern
Oscillation Index (pressure difference) between Darwin Australia
and Tahiti, increased rain in Peru and Ecuador, droughts in
northeast Brazil, Namibia and parts of Australia. The temporal
and spatial relationships among these and other events could be
studied to both understand and predict El Nino.
Assume a temporal data model where a map object is an
object made up
of time instants of maps. An event is then defined as a binary mask
over the map object. Note that the transformation is over the entire
map object, i.e., for each coordinate point and time instant. Thus,
an event is a spatial and temporal object. In addition attributes
may be added to the event. The attributes associated with an event
will embody remnants of the original data. For instance, the "net
primary production" would be an appropriate attribut...