Debates over directions of fluid flow in the presence of temperature gradients revolve around implications for contact metamorphic terranes. Studies of the Notch Peak contact-metamorphic aureole by Ferry and Dipple [1992] and Nabelek and Labotka [1993] focused the deliberations.
Ferry and Dipple [1992] proposed that H
O-rich fluids
that drove decarbonation reactions in the aureole flowed up temperature
from low-grade rocks toward the igneous pluton that was the source of
heat during metamorphism. Nabelek and Labotka [1993],
building on their previous studies and those of their co-workers,
presented evidence intended to controvert the model of Ferry and
Dipple. They argued that the flow of aqueous fluid was down
temperature away from the pluton.
Both proposed flow regimes are adequate to explain the major features of
the aureole: a sequence of diopside and wollastonite isograds in
calcareous argillites up temperature toward the pluton, and a general
decrease in whole-rock argillite carbonate 
values
toward the pluton. Although the distinction between the proposed models is
usually cast in terms of direction of flow relative to temperature, the
essence of the debate has as much to do with gradient versus fluid-driven
reactions and, by corollary, the origin of the infiltrated fluid. Ferry
and Dipple propose flow coupled with gradient reactions while Nabelek and
Labotka propose flow coupled with fluid-driven reactions.
Ferry and Dipple based their model on the assumption that fluid
entered the aureole in chemical and isotopic equilibrium with the
distal country rock. The reactions that ensued were thus envisaged to
have been gradient reactions. Reactant and product minerals are predicted
to coexist over a distance prescribed by the rock alteration index 
, and consumption of reactant phases, comprising
gradient reaction fronts, should not coincide with isograds if this
model is correct (isograd is used here to refer to the first
appearance of product minerals up temperature). Accordingly, Ferry
and Dipple focused on field relations in the zones between the
diopside and wollastonite isograds in which the relative positions of
lines that apparently demarcate loss of reactant minerals (gradient
fronts) relative to isograds serve to constrain their model. Flow
up-temperature toward the pluton reproduces these relations. They
interpreted the isotope data as being indicative of a gradual change
in 
with distance as is expected for gradient
isotopic exchange. Up-temperature flow was invoked to explain the
decrease in argillite 
toward the pluton.
Nabelek and Labotka presented trace element evidence that indicate fluid
that infiltrated the aureole came from the pluton and was not in
equilibrium with the country rock. Fluid-rock reactions were therefore
envisaged to be fluid-driven. Reactant and product minerals are not
expected to coexist over finite distances because reaction progress is
driven to completion by the difference in composition between infiltrated
and equilibrated fluid. Isograds are thus fluid-driven reaction fronts in
this model, and their
positions are controlled by 
and the difference in
composition between infiltrated and equilibrated fluid. Because fluid
compositions at equilibrium are a function of temperature, the positions
of the fronts envisaged by Nabelek and Labotka are a function of 
. In this circumstance the rock alteration index influences the
positions of both gradient and fluid-driven fronts, although in different
ways. Nabelek and Labotka interpreted the isotope data as defining a sharp
step-like change in 
with distance, consistent with a
fluid-driven exchange front.
The fundamental unknowns are the composition of the fluid relative to
rock and the nature of the fluid-rock reaction. Derivation of
reactive fluid from the pluton implies down-temperature flow,
and derivation from distal country rock implies up-temperature
flow assuming simple radial fluid stream lines, but there is little
inherent in the direction of flow relative to 
itself that
requires one model or the other. The salient features of the isotope
data, for example, including the existence of a front like that
described by Nabelek and Labotka, could be explained by
up-temperature flow of a nonequilibrium fluid (instead of the down-T
flow posited by the latter authors) if that fluid had
values several per mil greater than
indicated by equilibration with distal country rock [figure 2 of
Dipple and Ferry, 1992]. One concludes that the oxygen
isotope data
can be explained by flow either up or down temperature given suitable
initial 
for the infiltrated fluid. The most
conspicuous mineralogical features of the aureole are also explained
by either up or down-temperature flow, but only where the degree of
equilibration between fluid and rock is specified expressly. For
example, no one has suggested that fluid in equilibrium with
the country rock flowed down temperature in the aureole because such
a flow regime would have caused mineralogical reactions to progress
in directions opposite to those observed [e.g., Dipple and
Ferry, 1992; Ferry, 1991]. Down-temperature flow is
permitted for fluid-driven reactions but gradient reactions coupled with
down-temperature flow are precluded by first-order observations.
Both models for development of the Notch Peak aureole are based on continuum mechanics and were acknowledged by the protagonists to be at least partly schematic. In the future, the assumptions inherent in the continuum models can be tested at sub-REV scales. Such investigations should help sharpen our picture of aureole formation. Nabelek and Labotka alluded to this fact in their discussion of enhanced fracture flow in the wollastonite zone relative to the outer aureole.