 |
| Titel |
18,000 displacement vectors and 44 positions surveys of RFID tracers show a normal diffusion of the bedload in a proglacial stream (Bossons glacier, France) |
| VerfasserIn |
Hervé Guillon, Jean-Louis Mugnier, Jean-François Buoncristiani |
| Konferenz |
EGU General Assembly 2016
|
| Medientyp |
Artikel
|
| Sprache |
en
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 18 (2016) |
| Datensatznummer |
250128136
|
| Publikation (Nr.) |
 EGU/EGU2016-8089.pdf |
|
|
|
|
|
| Zusammenfassung |
| Bedload transport is a stochastic process during which each particle hops for a random length
then rests for a random duration. In recent years, this probabilistic approach was
investigated by theoretical models, numerical simulations and laboratory experiments.
These experiments are generally carried out on short time scales with sand, but
underline the diffusive behaviour of the bedload. Conversely, marked pebbles in natural
streams have mainly be used to infer about transport processes and transport time of
the bedload. In this study, the stochastic characteristics of bedload transport are
inferred from the radio-frequency identification (RFID) of pebbles. In particular, we
provide insights for answering the following question : is the bedload transport
sub-diffusive, normally diffusive or super-diffusive at the long time scale (i.e. global range)?
Experiments designed to investigate the phenomenology of bedload transport have been
carried out in the proglacial area of Bossons glacier. This 350 m long alluvial plain
exhibits daily flood from the glacial system and is still redistributing material from
catastrophic events pre-dating our investigations. From 2011 to 2014, the position of the ∼
1000 RFID tracers have been measured by a mobile antenna and a differential GPS
during 44 surveys providing ∼ 2500 tracer positions. Additionnaly, in 2014, 650 new
tracers were seeded upstream from a static RFID antenna located at the outlet of the
study area. For the 1 to 32 cm fraction surveyed, both mobile and static antenna
results show no evidence for a significant export outside of the surveyed zone. Initial
data have been maximized by using each possible campaign pairs leading to ∼700
campaign pairs and more than 18,000 displacement vectors. To our knowledge, this is
one of the most extensive dataset of tracers positions measured in a natural stream
using the RFID methodology. Using 152 campaigns pairs with at least 20 retrieved
tracers,r standard probability distributions were tested against the observed travel
distances. Regardless of the time scale, heavy- and light-tailed distributions provide a
convincing statistical description of measured data. No single distribution is significantly
better than the others. Conversely, the distribution of tracers positions in the system
and its time evolution is best described by the normal distribution. Its standard
deviation scales with time as σ ∝ t0.45±0.12 which suggests a nearly normal diffusive
behaviour. The measured virtual velocities and a simple probabilistic model using the
time evolution of the mean (i.e. drift) and standard deviation (i.e diffusion) show
that the mean bedload transfer time is greater than 5 years. RFID tracers appear
as a promising tool to investigate stochastic characteristics of bedload transport. |
|
|
|
|
|
|