Meaningfully Integrating Big Earth Science Data

Abstract:

After taking the technical hurdles to deal with big earth observation
data, large challenges remain to avoid that operations are carried
out that are not meaningful. Examples of this are summing things
that should not be summed, or interpolating phenomena that should
not be interpolated. We propose a description of data at the
level of their meaning, to allow for notifying data users when
meaningless operations are being executed. We present a prototypical
implementation in R.