Simon- I wouldn't call this a
definitive reference, but this treatment agrees with my
recollection from physical chemistry
http://www.edinformatics.com/math_science/suspensions_colloids.htm
The continuum is solution (= homogeneous mixture) - colloid -
suspension (=heterogeneous mixture) and at the extreme end,
separate phases rather than a mixture. An emulsion is a type of
colloid that involves only liquids. This distinction presumes a
particular "scale" is interest - the molecular scale. With this
assumption, and the criteria of the significance of Brownian
motion, I would say that Brownian motion is not, under any
conceivable scenario, a significant factor in the motion of
crocodiles and so crocs are neither dissolved nor colloidal
particles (not that you were suggesting they were, I just take the
extreme case to make a point). However as to particles in general
there is some fuzziness. There is not a sharp cut off where
Brownian motion changes from significant to insignificant.
Similarly there is not a sharp cutoff between suspended particles
and separated phases. As described here
(
http://en.wikipedia.org/wiki/Sediment_transport), particle
Reynolds numbers, dimensionless parameter dependent on
characteristics of the particle as well as the flow, can be used
to characterize the system. In borderline cases, judgement is
required to select the appropriate particle Reynolds number for
the intended purpose.
In all such cases of classification, there are multiple reasons to
question the observation. Accuracy of such observations can be
themselves estimated - higher accuracy for a classification based
on a more accurate estimate of a continuous property, lower
accuracy when the property estimate falls into the borderline zone
of the soft classifier. This is admittedly mixing uncertainty from
different kinds of sources (errors in observation with soft
classification) into a common "confidence" estimate - that's what
Bayesians do, and I'll confess to being an unrepentant Bayesian.
I believe an examination of soft classification from the
perspective of Bayesian statistics would be helpful to the
ontological community in getting a grip on the capabilities and
limitations of the hard classifications that appear to be built
into the less expressive KR languages. There is a large body of
work on this problem, and it is not a good use of our time to
reinvent this wheel.
Tara
On 10/18/13 10:37 PM,
Simon.Cox@xxxxxxxx wrote:
Maybe.
But it also depends on the size of the pipe and the rate of
flow.
The
reason that crocodiles are interesting is
(a)
In
this context they are not fundamentally different to any
other kind of suspended particle.
(b)
They
are a real water quality indicator, both in terms of the
capacity of a water body to support life, and the capacity
of the water body to support recreation!
(c)
They
are memorable!
It
seems that Torsten and Boyan are suggesting there is a
fundamental distinction between dissolved and suspended
contaminants, but I’m not sure there is a firm line. When I
did physical chemistry there was a continuum from
solution-suspension-emulsion-colloid, but in the case of
suspension I’m not sure if there is a critical particle size
– look at the movies from the Fukushima tsunami to see how
very large objects can be part of the flow.
You
also appear to be accepting that hydrodynamics is the
use-case under discussion. There are other cases in which
water contaminants play a role.
Simon
I
think it's just more simple: Torsen gave us elements
of hydrology theory (formal one;-) abstracted from
bio-objects.
Suppose
this theory can predict flood movement and other
hydrology processes.
But
where crocodiles? They are neglected. Abstraction.
They
do not impact to hydrological processes of interest.
Do they?
I
think they have an amendment to the coefficient of
viscosity of the water in this case;-)
_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontology-based-standards/
Subscribe: mailto:ontology-based-standards-join@xxxxxxxxxxxxxxxx
Config/Unsubscribe: http://ontolog.cim3.net/mailman/listinfo/ontology-based-standards/
Shared Files: http://ontolog.cim3.net/file/work/OntologyBasedStandards/
Wiki: http://ontolog.cim3.net/cgi-bin/wiki.pl?OntologyBasedStandards