Quantifying erosion in mountainous landscapes Mikaël ATTAL Marsyandi valley, Himalayas, Nepal Acknowledgements: Jérôme Lavé, Peter van der Beek and other.

Quantifying erosion Quantifying erosion in mountainous in mountainous

landscapeslandscapes

Mikaël ATTALMikaël ATTAL

Marsyandi valley, Himalayas, Nepal

Acknowledgements: Acknowledgements: Jérôme Lavé, Peter van Jérôme Lavé, Peter van der Beeder Beek and other scientists from LGCA k and other scientists from LGCA

(Grenoble) and CRPG (Nancy)(Grenoble) and CRPG (Nancy)

Eroding landscapes: Eroding landscapes: fluvial processesfluvial processes

National Student Satisfaction survey

(4th year)

Lecture overviewLecture overview

I. Bedrock erosion processesI. Bedrock erosion processes

II. Quantifying fluvial (and landscII. Quantifying fluvial (and landscape)ape) erosion erosion on the long-term on the long-term

IIIII. Quantifying fluvial erosion on the short-termI. Quantifying fluvial erosion on the short-term

I. Bedrock erosion processes

Abrasion (bedload impact)

Abrasion (suspended load)

Plucking (Ukak River, Alaska, Whipple et al., 2000)

Cavitation (www.irrigationcraft.com)

Abrasion (bedload impact)

Amount of abrasion is a function of: kinetic energy = 0.5mv2; angle of impact; difference in rock resistance between projectile and target


Plucking

BEDLOAD EXERTS A KEY ROLE

Amount of erosion is a function of: joint density; stream power; kinetic energy of impacts = 0.5mv2; angle of impact.

Whipple et al., 2000 (Ukak River,

Alaska)


Requires turbulence (eddies) affects mostly obstacles protruding in the channel (e.g. boulders)

Abrasion by suspended load

Whipple et al., 2000


Most of the time, sediment is resting on the bed and protects it from erosion bedrock erosion (abrasion by bedload impacts + plucking) happens during floods

V

D



V

D

Consider 1 point in the channel, at a given time, during 1 flow event

Because sediments in river include a wide range of grain sizes, some

particles will move while some others (larger) will rest on the river bed

TOOLS & COVER

Transport capacity: Qc = k(τ – τc)3/2

where k and τc are constants [Meyer-Peter-Mueller, 1948]

Fluvial shear stress: τ0 = ρ g R S

Stream power per unit length: Ω = ρ g Q S

Whipple et al., 2000: process-based theoretical analysis within the frame of the SPL

e = KAmSn = kτa where n = 2a/3 (and m is adjusted to obtain m/n = 0.5)

Remark:

if n = 1, m = 0.5 and a = 3/2 Incision Specific Stream power (law 2).

if n = 2/3, m = 1/3 and a = 1 Incision basal shear stress (law 3).

Abrasion (bedload)Abrasion (bedload) Not analyzedNot analyzed Not analyzedNot analyzed

Abrasion (suspension)Abrasion (suspension) nn = 5/3 = 5/3 aa = 5/2 = 5/2

PluckingPlucking nn = 2/3 = 2/3 11 aa = 1 = 1 3/2 3/2

CavitationCavitation nn up to 7/3 up to 7/3 aa up to 7/2 up to 7/2


II. Quantifying fluvial (and landscape) erosion on the long-term (103-106 years)

1) Fluvial erosion rates using terrace dating

2) Catchment-wide erosion rates using the fluvial network as an “age homogenizer”

“Long-term” fluvial erosion rates (103-106 years): fluvial terraces

Courtesy J. Lavé

STRATH TERRACES

Note: rivers can erode and form terraces even without uplift

Strath terraces: thin (or no) alluvium cover, contact alluvium-bedrock relatively flat

Siwaliks hills, Himalayas (J. Lavé)

Central Range, Taiwan

Siwaliks hills, Himalayas (J. Lavé)

Strath terraces: thin (or no) alluvium cover, contact alluvium-bedrock relatively flat

Fluvial incision rates using strath terrace dating

Age = 0 yr

Age = n yr

h Incision rate = h/n

Dating methods:

- 14C,

- Optically stimulated luminescence (OSL),

- Cosmogenic nuclides.


Bagmati River, Himalayas (Lavé & Avouac, 2000, 2001)


Terraces are correlated in the field + using remote sensing




Relatively constant incision rates since the end of the Pleistocene (PL3 is ~ 22ky old)

Reminder: the Quaternary Period includes the following epochs: Pleistocene (1.8 Ma ~12 ka) and Holocene (~12ka present)

FILL TERRACES: usually the result of landslides damming the valley (or large alluviation events filling narrow valleys)

Tal, Marsyandi valley, Himalayas

Upstream of the dam

!

FILL TERRACES: usually the result of landslides damming the valley (or large alluviation events filling narrow valleys)!

Chame, Marsyandi valley, Himalayas

Thick alluvium, up to hundreds of meters, contact alluvium- bedrock highly irregular.

Local effect must not be used to determine long-term erosion rates.

image28.webshots.com

The events that lead to the formation of fill terraces are relatively frequent in actively eroding landscapes

Amount of erosion at a given point along the river

Time (x 105 years)

Models, long-term measurements

Reality

FILL TERRACES: usually the result of landslides damming the valley (or large alluviation events filling narrow valleys)!

Terrace dating methods

a) 14C on organic debris in alluvium (up to ~40 ka).

3 carbon isotopes: 12C (natural abundance 98.89 %), 13C (n.a. 1.11 %) and 14C (n.a. 1 part / trillion).

[14C] in the atmosphere is ~ constant (equilibrium between rate of production and decay) and is ~ to [14C] in living organisms.

14C formed in the atmosphere (interaction between cosmic rays and N molecules):14N + n 14C + p

When organism dies no more exchange with atmosphere the number of 14C atoms decreases due to radioactive decay.

14C (or radiocarbon) is a radioactive isotope which decays with a half-period of 5730 years. Age of terraces can be

estimated by counting the number of 14C atoms in organic fragments (assuming that the time between the organism’s death and its incorporation into the alluvium is negligible).

http://www.irb.hr/en/str/zef/z3labs/lna/C14/

14C

orga

nis

m /

14C

atm (

%)


b) Optically stimulated luminescence: burial ages of quartz or feldspar crystals, ages from 100 yrs to 350 000 yrs.

Radioactive isotopes + cosmic rays charge carriers (e.g., electrons e-, electron holes h+) travelling in crystals

http://www.ndt-ed.orghttp://www.enigmatic-consulting.com

Charge carriers

Charge carriers can become trapped in lattice defects. They progressively accumulate in these “traps” over geological timescales.



Radioactive isotopes + cosmic rays charge carriers (e.g., electrons e-, electron holes h+) travelling in crystals

http://www.ndt-ed.orghttp://www.enigmatic-consulting.com

Charge carriers

Charge carriers can become trapped in lattice defects. They progressively accumulate in these “traps” over geological timescales.

Exposure to light, heat, or high pressures can release charge carriers from trapping sites reset the system

The release process is associated with a photon release. Number of photons released = f (number of trapped charge carriers released).



Sunlight releases trapped charge carriers.

If a crystal gets buried, charge carriers are going to accumulate in trapping sites.

The longer the burial, the larger the number of trapped charge carriers.

http://suppelab.gl.ntu.edu.tw

Optical stimulation (light)

release of charge carriers release of photons light emission

The older the terrace, the longer the burial, the higher the number of trapped charge carriers the larger the number of photons released with the charge carriers the higher the intensity of the light emitted!


c) Cosmogenic Nuclides: exposure ages.

Cosmic rays interact with atoms in the atmosphere and in the rocks exposed at the surface of the Earth nuclear reactions cosmogenic nuclides.

Examples: 3He, 10Be, 14C, 21Ne, 26Al, 36Cl.

Cosmogenic nuclides accumulate in minerals in the 1-2 m thick layer at the top of the Earth.

Stable

T1/2 = 1.5 Ma

T1/2 = 5730 a

Stable

T1/2 = 0.73 Ma

T1/2 = 0.3 Ma

The longer the rock exposure, the higher the amount of cosmogenic nuclides

Depth

Cosmogenic nuclide production rate

1-2 m

Concentration in cosmogenic nuclides in minerals = f (EXPOSURE TIME, latitude, altitude, topography, type of mineral, type of cosmogenic nuclide).


c) Cosmogenic Nuclides: exposure ages.

Beryllium: 9Be = stable isotope; 10Be = cosmogenic isotope formed by interactions between cosmic rays and O, N, Si, Mg, Fe. Beryllium in Quartz frequently used in geomorphology to date objects up to millions of years old.

Chlorine: 35Cl and 37Cl = stable isotopes; 36Cl = cosmogenic isotope formed by interactions between cosmic rays and Ar, Fe, K, Ca, Cl. Chlorine in calcite is a method which begins to be reliable to date objects up to millions of years old.

http://web.ges.gla.ac.uk/~jjansen

Boulders on terraces

Bedrock strath terrace© Scott T. Smith/CORBIS

http://www.futura-sciences.com/

II. Quantifying fluvial (and landscape) erosion on the long-term (103-106 years)

1) Fluvial erosion rates using terrace dating

2) Catchment-wide erosion rates using the fluvial network as an “age homogenizer”

“Detrital methods”

Assumption: time spent in the fluvial network is negligible

Photo Eric Gayer

Catchment-wide erosion rates

a) Cosmogenic ages on fluvial sands (Q + Grt). Courtesy Eric Gayer


Main limitation: assumption that landscape is eroding at a constant rate through time

Uplift

Erosion1-2 m

1-2 m

a) Cosmogenic ages on fluvial sands (Q + Grt).



Uplift

Erosion1-2 m

1-2 m




Uplift

Erosion1-2 m

1-2 m




Uplift

Erosion1-2 m

1-2 m




Uplift

Erosion1-2 m

1-2 m




Uplift

Erosion1-2 m

1-2 m

Landslide

gives the impression that the catchment includes zones with low, moderate and extremely high erosion rates!



b) Detrital termochronology: fission tracks

Bernet & Garver, 2005

pangea.stanford.edu

Fission tracks in zircon or apatite




If erosion rate is constant, lag time is constant.

Example: lag-time = 20 Ma

Deposition age (age of sediment td) (Ma)

FT age (tc)

0 10 20 30 40 50 Ma50

40

30

20

10

tc = 50 Ma

td = 30 Ma0







FT age (tc)

0 10 20 30 40 50 Ma50

40

30

20

10

tc = 40 Ma

td = 20 Ma0







FT age (tc)

0 10 20 30 40 50 Ma50

40

30

20

10

tc = 30 Ma

td = 10 Ma0







FT age (tc)

0 10 20 30 40 50 Ma50

40

30

20

10

tc = 20 Ma

td = 0 Ma0

Slope 1:1

Short lag-time

Long lag-time







FT age (tc)

0 10 20 30 40 50 Ma50

40

30

20

10

tc = 30 Ma

td = 15 Ma0

Let’s imagine that erosion rate increases at 30 Ma Lag-time = 15 Ma







FT age (tc)

0 10 20 30 40 50 Ma50

40

30

20

10

tc = 20 Ma

td = 5 Ma0

Let’s imagine that erosion rate increases at 30 Ma Lag-time = 15 Ma


b) Detrital termochronology: Ar/Ar or K/Ar methods (very simplified here)

39K is stable.40K decays into 40Ar (gas) with a half-life of 1.25 billion years.

Degassed at high temperature, accumulates in minerals at temperatures < closure temperature

Isotherm corresponding to closure temperature

T > closure temperature: 40Ar degassed

T = closure temperature clock starts

40Ar accumulates in mineral. Amount of 40Ar = f (time since crossing the isotherm)

Biotite: 300 ºCMuscovite: 400 ºCHornblende: 550 ºC


b) Detrital termochronology: Ar/Ar or K/Ar methods (very simplified here)

Central Himalayas, Nepal (Wobus et al., 2005)

Ar/Ar ages on detrital muscovite

Isotherm 400 ºC Migration of the MCT?

Quantifying erosion in mountainous landscapes Mikaël ATTAL Marsyandi valley, Himalayas, Nepal Acknowledgements: Jérôme Lavé, Peter van der Beek and other.

Documents

quantifying erosion

flat slide

longterm erosion rates

taiwan slide

shortterm slide

uplift slide

landscape erosion

fluvial incision rates