Travertine terracing: patterns and mechanisms - Tufas and Speleothems: Unravelling the Microbial and Physical Controls - page 339

Travertine terracing: patterns and mechanisms

ØYVIND HAMMER*, DAG K. DYSTHE & BJØRN JAMTVEIT
Physics of Geological Processes, University of Oslo, PO Box 1048 Blindern, 0316 Oslo, Norway
*Corresponding author (e-mail: [e-mail])

Abstract: Travertine terracing is one of the most eye-catching phenomena in limestone caves and
around hydrothermal springs, but remains fairly poorly understood. The interactions between water
chemistry, precipitation kinetics, topography, hydrodynamics, carbon dioxide degassing, biology,
erosion and sedimentation constitute a complex, dynamic pattern formation process. The processes
can be described and modeled at a range of abstraction levels. At the detailed level concerning the
physical and chemical mechanisms responsible for precipitation localization at rims, a single
explanation is probably insufficient. Instead, a multitude of effects are likely to contribute, of
varying importance depending on scale, flux and other parameters.

Travertine terracing is undoubtedly among the
most spectacular geological phenomena on Earth.
Ranging in vertical scale from millimetres to tens
of metres, travertine terraces form intricate, delicate
patterns as well as imposing waterfalls. Such ter-
races are common not only in limestone caves and
around hot springs, but also in streams and rivers
in limestone terrain (Pentecost 2005, 59 - 66). In
addition, analogous patterns are found in completely
different systems, such as silica sinter deposits and
water ice. Some of the more spectacular examples
of travertine terracing are found in Yellowstone
National Park (Bargar 1978; Fouke et al. 2000),
Pamukkale in Turkey (Altunel & Hancock 1993),
Huanglong Scenic District in China (Lu et al.
2000) and northern Spitsbergen, Norway (Hammer
et al. 2005). A list of travertine occurrences in
Europe and Asia Minor with notes on terracing
was given by Pentecost (1995). Travertine terraces
formed an inspiration for Renaissance ornamental
water cascades (Berger 1974). However, in spite
of their great interest, both aesthetically and scienti-
fically, the formation of travertine terraces has until
recently received only scattered scientific attention,
especially on the theoretical side (Wooding 1991
provides one notable exception).
When trying to understand this pattern formation
system, a series of fundamental questions arise.
What processes are responsible for the enhanced
precipitation at the terrace rim? On a higher level
of abstraction, how do these local processes lead
to global self-organization? How do the terraces
evolve in time, and what are their statistical proper-
ties under different conditions? Answering such
questions requires a cross-disciplinary approach,
involving field work, laboratory experiments, com-
puter modeling and mathematical theory in order
to study the complicated feedbacks between hydro-
dynamics,

precipitation and possibly particle transport and
biology.

Morphology and hydrodynamics

The terminology of travertine terrace morphology is
confusing. Speleologists often use the roughly
equivalent terms 'rimstone' and 'gours', with
'microgours' referring to cm-scale terraces. For
open-air travertine systems, morphological classifi-
cation schemes have been proposed by several
authors. Pentecost & Viles (1994) define 'barrages'
as terraces that are filled with water, forming pools
and lakes. 'Cascades' are smoother forms on steep
slopes, often with smaller terrace-like structures
superimposed on them (Fig. 1). Pentecost (2005)
uses 'dams' instead of barrages. Fouke et al.
(2000) and Bargar (1978) use three size categories
of barrages, namely 'terraces' with areas of tens of
square metres, 'terracettes' of a few square metres
and 'microterracettes' of a few square centimetres
or less (Fig. 2). Microterracettes are therefore analo-
gous to microgours. Pentecost (2005) suggests the
term 'minidam' for a dam with an interdam distance
(IDD) ranging from 1 cm to 1 m.
In this paper we will use the classification of
Fouke et al. (2000), extending it to the analogous
speleothems. In addition, we recognize that on
steep slopes, microterracettes will not form pools
(and are therefore not barrages) but grade into
subdued 'microridges' normal to flow. We also
use 'terrace' as a general term regardless of size.
We use 'rim' for the top of the outer wall of the
terrace, and 'pool' for the water body behind it.
The rim is usually convex outwards (downstream),
or consists of a series of near-parabolic lobes point-
ing in the downstream direction (Fig. 2). Inside
pools, flow velocity is small due to the larger water chemistry, calcium carbonate