using isotope techniques to track terrestrial sediments ... · using isotope techniques to track...
TRANSCRIPT
Using isotope techniques to track terrestrial sediments from soil sources
to freshwater systems
Christine AlewellEnvironmental Geosciences, University of Basel, Switzerland
21.3.2017 JRC, ISPRA
This talk is on how to assess……
Soil erosion Sediment input to freshwaters
On site effects of erosionSoil fertility and productivitybiodiversitycarbon storageSoil stability
Off site damage of erosionFreshwater deteriorationEutrophication (P, N)Clogging of river bedsImpact to infrastructure
with isotope techniques
Contents
Tracking erosion on site with stable isotopes
Quantifying erosion on-site with radionuclides
Off site sediment tracking with CSSI markers
C3 - plants
δ13CVegetation ~ - 28.6 ± 0.9 ‰(range -22 bis – 32 ‰)
δ13CHumus > δ13CVegetation
δ13CMin.Soil > δ13CHumus
12CO2
12CO2
- 28.6 ‰
- 23.0 ‰
Photos: Marco Walser, WSL
Tracking erosion on site with stable isotopes
0δ13C
dept
h (c
m)
60
-30 -20
Isotope depth profiles in undisturbed soils:
vegetation signal
δ13CVeg = -28.6 ± 0.9 ‰
δ13C % carbon in soil
%C δ13C %
C
Two Swiss Alpine Sites
Lake Soyang Watershed, Korea
Punch Bowl Watershed
Stable Isotopes as indicators of erosion: Lake Soyang watershed
Carbon content vs. δ13C for reference sites and erosion transectsMeusburger et al., 2013 Biogeosciences 10
? Quantification of on-site erosion?
Assessment of 137Cs and 239+240Pu as soil erosion tracer in alpine grasslands
Stable isotopes as qualitative indicators of soil disturbance
Validation of reference sites for quantification with FRN
One decade of FRN wet deposition1950ties and 1960ties
Quantification of Soil Erosion with Fallout radionuclides (FRN)
erosion:depletion in FRN
sedimentation:increase in
FRN
reference site
Quantification of Soil Erosion with Fallout radionuclides (FRN)
Advantages:• global distribution• retrospective assessment• spatially distributed data• only one sampling campaign
required • both erosion and deposition
Limitations:• choice of reliable reference
sites• specific detection systems
Most commonly used FRN: 137Cs
26.4.1986Cs-137 deposition
by a few single rain events
Complicating life in Europe: 137Cs fallout by Chernobyl
Pitfalls of Cesium in EuropeHuge spatial heterogeneity of “reference” sites
Input from Chernobyl in April 1986Overall disturbance of
alpine slopes?
Check with stable isotopes
Ursern Valley (Uri) Val Piora (Ticino)(Bq.kg-1) n=6 n=6 n=9 n=7, without 2
outliersMean 131 106 420 142Stdev 17 43 475 24CV (%) 10 40 110 20
Konz et al. (2010)
Polek (2011) (Juretzko, 2011)
Suitability of 239+240Pu versus 137Cs
0
20
40
60
80
100
120
140
160
0
2000
4000
6000
8000
10000
12000
14000
16000
18000
1 2 3 4 5 6
Pu-In
vent
ory
[Bq
m-2
]
Cs-In
vent
ory
[Bq
m-2
]
Reference sites Ursern Valley
Cs-137 Pu-239+240
0
20
40
60
80
100
120
0
5000
10000
15000
20000
25000
30000
1 2 3 5 6 7 9
Pu-In
vent
ory
[Bq
m-2
]
Cs-In
vent
ory
[Bq
m-2
]
Reference site Piora Valley
Cs-137 Pu-239+240
Alewell et al., 2014, Chemosphere
Urseren Pioran 6 7
137Cs (Bq m-2) mean 6892 10355CV (%) 32 98
239+240Pu (Bq m-2) Mean 83 77CV (%) 13 17
And then there is 210Pbex
Mabit et al., 2014
Advantages: Fallout is continuous over time Long term (100 years)
Disadvantages: No direct measurement! high uncertainty related to
determination 210Pbex especially if total 210Pb is close to supported 210Pb
Quite often below detection limit
Multi-Radionuclide Approaches Validation of FRN against each other (South Korea, no Chernobyl input and prior to Fukushima, Meusburger et al., 2016 STOTEN):
Same order of magnitude for 239+240Pu versus 137Cs (difference in conversionmodels, difference in diffusion and migration coefficients)
No comparability to 210Pbex
Multi-Isotope Approaches
Determine origin of FRN 240Pu/239Pu (238Pu/239Pu and 241Pu/239Pu with limitations): nuclear weapon production vs weapon fallout vs accidental and routine releases from nuclear installations 137Cs/239+240Pu or 134Cs/ 137Cs: 137Cs input from Chernobyl
Assessment of particle size correction factor preferential transport of 137Cs compared to 239+240Pu
(Meusburger et al., 2016 STOTEN)
Advantages of 239+240Pu as soil erosion tracer not influenced by nuclear accident fallout (with the exception of regions
with close proximity to accident sites): relatively low heterogeneity of reference sites
analytical advantage of lower cost and higher sample throughput either with ICP-MS or alpha spectrometry compared to both 137Cs or 210Pbex
main isotopes (239Pu and 240Pu) have long half-lives of 24’110 and 6’561 years: advantage in terms of its long term use
seems to be a much more reliable soil redistribution tracer compared to 210Pbex
Medium to short term soil erosion rates might be assessed via re-sampling techniques (e.g., Porto et al., 2014)
New conversion model MODERN available, to convert Pu inventory changes into soil redistribution rates
t ha2 yr-1 mm yr-1 Specifications Soil degradation
Konz et al. (2010) Alewell et al. (2014)
30* 14*
3 1.4
137Cs measurements; hot spots 239+240Pu measurements; hot spots
Meusburger et al. (2010) 16* 1.2*
1.6 0.12
sheet erosion modelled hot spots sheet erosion modelled average
Meusburger and Alewell (2009) 0.60* 0.06 Landslides, measurements
Meusburger et al. (2009, 2010) 1.8* 0.18 Average sum erosion +land slides Geomorphological rate of Alpine soil formation depending on age
10 - 18 ky 0.3-0.6* 0.03-0.06 old surfaces
1 - 10 ky 0.6-3.5* 0.06-0.35 younger surfaces
0 - 1 ky 3.5-20* 0.35-2.0 Very young or strongly eroded sites
Significance of erosion rates in our alpine catchment?
Conversions from t km2 yr-1 to mm yr-1 with soil bulk densities of 1 t m-3 (surface horizon)Alewell, Egli, Meusburger, 2014; JSS
>> 1 t ha-2yr-1 Non sustainable soil use: degradation of soils Problem will most likely increase in the future (climate change, land use change)
Sediment/ Organic matter Source Attribution
Forest
Pasture
Arable field
d13C
d13C %?
%?
%?Catchmenterosion
Soil
Catchment source areas
Concept
Soil/SedimentSample
Dry weight, %C, %N
Bulk δ13C, δ15N
Extraction Total lipid extract Separation Acid fraction
Fatty acid methyl esters
(FAME)
GC-MS
GC-FID
GC-IRMS
Catchment Enziwigger
Methods and Implementation
sediment basket
Suspended Sediment sampler
Sampling fall and winter 2009/2010 and 2010/2011
SchindlerWildhaber et al., HESS 2012
...all to often..... ...not only.....
Sampling fall and winter 2009/2010 and 2010/2011
Sediment source attribution with CSSI
Ale
wel
l, B
irkho
lz, M
eusb
urge
r et a
l., 2
016
δ13C of FAs C26:0 versus C28:0 of sediment sources and suspended sediments (SS) at the three sites (A, B and C) in the Enziwigger catchment.
Sediment source attribution to the Enziwigger
Contribution of the different sediment source areas to the SS for two or three sources. BF = base-flow event, HF = high-flow event.
% sediment contribution from 2 Tracers/3 Sources (IsoSource)Site Event Forest Agriculture %Forest % Pasture % Arable A BF 70 30A HF 2010 85 15A HF 2009 60 40B BF 37 63 28.2* 16.6* 55.2*B HF 2010 94 7 92.1 2.4 5.5B HF 2009 78 22 69.5 9.4 21.1C BF 34 66 31.8 23.6 44.6C HF 2010 72 29 64.7 12.3 23.0C HF 2009 55 45 49.2 17.7 33.1
Conclusions
Stable isotopes can be used to qualitatively track soil erosion on-site and validate reference sites for FRN based erosion assessment
Quantification of soil erosion assessment with FRN:239+240Pu seems more reliable than 137Cs in Europe due to Chernobyl input and generally more reliable than 210Pbex
Measurement of FRN: Analytical advantages of 239+240Pu (ICP-MS)
Off site tracking with CSSI: high effort in analytical devices, investment and lab staff, but strong tool to track down sources of sediments: plenty more to explore
Thank you for your attention!
KatrinMeusburger
Lionel MabitNow at IAEA, Vienna Mike Ketterer
University of Denver
Ji-Hyung ParkEwha Womans University,
South Korea
Axel Birkholz
Yael Schindler Now at FOEN, Switzerland
Laura Arata
Markus ZehringerKantonslabor, Basel
Funding:Swiss National Science FoundationSwiss Federal Office of the EnvironmentIAEA, Vienna
Discussion
Isotopes as indicators of soil degradation in the Swiss Alps
Upland soils with no visible erosion
r > |-0.85 |
Soils prone to erosion r ≤ |-0.80|
Schaub and Alewell, 2009, Rap Comm Mass Spec 23
Val Piora (Ticino) Ursern Valley (Uri)(Bq.kg-1) n=9 without 2
outliersn=6 n=6
Mean 420 142 106 131Stdev 475 24 43 17CV (%) 110 20 40 10
(Juretzko, 2011) Polek (2011) Konz et al. (2010)
Validating disturbance of reference sites with stable isotopes
Brun, 2012
Conversion: MOdelling Deposition and Erosion rates with RadioNuclides (MODERN)
32
The MODERN code is available atmodern.umweltgeo.unibas.ch
Arata et al., 2016 a, b; J. Environ. Radioact.
Results of MODERN
33
y = 2.02x - 0.78R² = 0.90
-40
-30
-20
-10
0
10
20
30
-20 -10 0 10 20
IM (t
ha-
1yr
-1)
MODERN (t ha-1 yr-1)
239+240Pu based estimates
y = 0.92x - 0.16R² = 0.99
-8-6-4-202468
10
-10 -5 0 5 10 15
MO
DER
N (t
ha-
1 yr
-1)
PDM (t ha-1 yr-1)
137Cs based estimates
y = 0.81x - 3.97R² = 0.63
-60
-40
-20
0
20
40
-60 -40 -20 0 20 40
MO
DER
N (t
ha-
1 yr
-1)
MBM2 (t ha-1 yr-1)
137Cs based estimates
y = 1.17x - 0.25R² = 0.65
-50
-40
-30
-20
-10
0
10
20
-40 -20 0 20IM
(t h
a-1
yr-1
)MODERN (t ha-1 yr-1)
239+240Pu based estimates
Meusburger et al., 2016; STOTEN
137Cs
239+240Pu
PDM: Profile distribution model; MBM2: Mass balance model 2 (Walling et al., 2002, 2014); IM: Inventory model (Lal et al., 2013)
=>transition of land use not considered yet
uncultivated cultivated
uncultivated cultivated
Sediment source attribution with CSSI
Alew
ell,
Birk
holz
, Meu
sbur
ger e
t al.,
sub
mitt
ed
δ13C of the FAs C26:0 and C28:0 at site A: Considering measurement un-precision, δ13C were corrected to the mixing line with linear regression
Sediment source attribution with CSSI
Ale
wel
l, B
irkho
lz, M
eusb
urge
r et a
l., su
bmitt
ed
δ13C isotopic signatures of FAs C26:0 versus C14:0 of sediment sources and suspended sediments (SS) at the three sites in the Enziwigger catchment.
Quantitative sediment source attribution
A system with n tracers is solvable for n+1 sources because you have n+1 equations:
fA + fB + fC = 1 f=fraction
𝝏𝝏𝟏𝟏𝟏𝟏𝑵𝑵𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔 = 𝒇𝒇𝒇𝒇 𝝏𝝏𝟏𝟏𝟏𝟏𝑵𝑵𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒇𝒇 + 𝒇𝒇𝒇𝒇 𝝏𝝏𝟏𝟏𝟏𝟏𝑵𝑵𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒇𝒇 + 𝒇𝒇𝒇𝒇 𝝏𝝏𝟏𝟏𝟏𝟏𝑵𝑵𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒇𝒇
𝝏𝝏𝟏𝟏𝟑𝟑𝒇𝒇𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔 = 𝒇𝒇𝒇𝒇 𝝏𝝏𝟏𝟏𝟑𝟑𝒇𝒇𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒇𝒇 + 𝒇𝒇𝒇𝒇 𝝏𝝏𝟏𝟏𝟑𝟑𝒇𝒇𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒇𝒇 + 𝒇𝒇𝒇𝒇 𝝏𝝏𝟏𝟏𝟑𝟑𝒇𝒇𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒇𝒇
If sources > n+1: mixing model approaches, calculation of all possible solutions within a range of specified uncertainty(e.g. model ISOSOURCE by Phillips and Gregg (2001))
-3-2.5
-2-1.5
-1-0.5
00.5
11.5
2
-36 -31 -26 -21 -16
δ15 N
δ13C
PastureForestArableSuspended Sediments
36
Swiss Alpine sites t km2 yr-1 mm yr-1
Konz et al. (2010)Alewell et al. (2014)
3000*1400*
31.4
137Cs based; hot spots239+240Pu based; hot spots
Meusburger et al. (2010) 1600*118*
1.60.118
sheet erosion modelled hot spotssheet erosion modelled average
Meusburger and Alewell (2009) 60* 0.060 Landslides, measurements Meusburger et al. (2009, 2010) 178* 0.178 Average sum
Literature dataDosseto et al. (2011) 400 0.400# lower range for Alps
20000 20# higher range for AlpsFelix and Johannes (1995) 440* 0.44 calcareous Alps, BavariaFrankenberg et al. (1995) 3000* 3# Flysch, Molasse, Allgäuer AlpsAmmer et al. (1995) 200 - 900* 0.2 - 0.9 Flysch, calcareous AlpsDescroix et al. (2003) 1400 -3300* 1.4 - 3.3 French AlpsIsselin-Nondedeu and Bedecarats (2007)
6000 - 18000 6 - 18# heavy rain events, French Alps
Comparison to other studies
*# gives the original published number which was converted (t km-2 yr-1 versus mm yr-1).
137Cs repeated sampling approach
Spatial reference
instead
Temporal reference:
Compare 2007 Cesium measurements with 2012
Re-sampling approach
Years ≠Sites =
Years =Sites ≠
Classical137Cs method