2011_Feb02_AM_Butman_278

coriandercultureΜηχανική

22 Φεβ 2014 (πριν από 3 χρόνια και 8 μήνες)

83 εμφανίσεις

EMERGING PERSPECTIVES ON
CONTINENTAL
-
SCALE

RIVERINE CARBON FLUXES


David Butman, Yale University

Edward G. Stets, U.S. Geological Survey

Peter Raymond, Yale University

Robert G. Striegl, U.S. Geological Survey

Acknowledgements
-


NASA Earth and Space Science Fellowship


NSF
-
Ecosystems CAREER


Yale Center for Earth Observation


Christopher Zappa, Tom Bott, Jody Potter, Patrick
Mulholland and William McDowell
(Gas Transfer
Meta
-
Analysis)
, Charlie Crawford, Kathleen Johnson,
Cory McDonald.


AQUATIC CARBON CYCLING

CO
2

IC


OC

Stream flow

CO
2

IC


OC

Upstream

Terrestrial

Hyporheic

Downstream

or

Coastal

IC


Inorganic carbon

OC


Organic carbon

AQUATIC CARBON CYCLING

CO
2

CO
2

CO
2

IC


OC

Stream flow

CO
2

IC


OC

IC

OC

CO
2

CO
2

Upstream

Terrestrial

Hyporheic

Downstream

or

Coastal

IC


Inorganic carbon

OC


Organic carbon

Sedimentation

AQUATIC CARBON CYCLING

CO
2

CO
2

CO
2

IC


OC

Stream flow

CO
2

IC


OC

IC

OC

CO
2

CO
2

Upstream

Terrestrial

Hyporheic

Downstream

or

Coastal

IC


Inorganic carbon

OC


Organic carbon

Sedimentation

AQUATIC CARBON CYCLING

CO
2

IC


OC

Stream flow

CO
2

IC


OC

Upstream

Terrestrial

Hyporheic

Downstream

or

Coastal

IC


Inorganic carbon

OC


Organic carbon

CO
2

“CO
2

Flux”

“Lateral
Flux”

Sedimentation

Water quality and streamflow data



USGS National Water Information System (NWIS).


National Stream Quality Accounting Network (NASQAN).


DATA SOURCES & TERMINOLOGY

Carbon fractions



Inorganic carbon (IC) and CO
2


Calculated from Alkalinity, temperature, pH


Assume alkalinity arises from
Σ
CO
2


Assume particulate fraction of alkalinity is negligible


Organic carbon (OC).


Total organic carbon (TOC)


unfiltered


Dissolved organic carbon (DOC)
-

filtered


Inorganic carb

(g C m
-
2

yr
-
1
)

>

4



<

No

LATERAL CARBON FLUXES

CO
2

FLUX FROM STREAMS


otal


lux:


.7 Tg C yr
-
1

Tota


(Tg


IC


TOC

CO
2


RIVERINE C FLUX IN PERSPECTIVE

USGS Load Estimator (LOADEST)
-


1.
Calibration data.



2.
Multiple regression.



3.
Calculate daily loads.



4.
Aggregate into annual fluxes.

Date

Stream
Flow (cfs)

TOC
(mg L
-
1
)

04/24/2008

17,100

5.04

05/20/2008

12,500

4.45

06/11/2008

5,390

4.69

07/10/2008

6,520

6.23

08/05/2008

6,520

4.69

08/20/2008

8,290

6.87

09/17/2008

10,600

7.31

10/14/2008

8,270

6.2

12/23/2008

37,300

7.7

01/15/2009

25,600

5.44

02/05/2009

13,300

3.92

02/17/2009

13,500

16.99

FLUX CALCULATIONS

USGS Load Estimator (LOADEST)
-


1.
Calibration data.



2.
Multiple regression.

)
cos(
)
sin(
ln
ln
ln
4
3
2
2
1
0
time
a
time
a
Q
a
Q
a
a
LOAD





3.
Calculate daily loads.


4.
Aggregate into annual fluxes.

Relationship between
load and stream flow

Seasonal variability

FLUX CALCULATIONS

USGS Load Estimator (LOADEST)
-


1.
Calibration data.



2.
Multiple regression.



3.
Calculate daily loads.



4.
Aggregate into annual fluxes.

Daily TOC loads (x10
6

g d
-
1
)

Jan-08
Jul-08
Jan-09
Jul-09
0
2000
4000
6000
FLUX CALCULATIONS

USGS Load Estimator (LOADEST)
-


1.
Calibration data.



2.
Multiple regression.



3.
Calculate daily loads.



4.
Aggregate into annual fluxes.

2008
2009
0.0
0.1
0.2
Annual TOC loads (Tg C yr
-
1
)

FLUX CALCULATIONS

1950
1970
1990
2010
0
200
400
600
800
1000
TOC (
n
= 5,126; 151 sites)

IC (
n

= 4,552; 161 sites)

DOC (
n

= 3,121; 144 sites)

DATABASE FEATURES

Observations

per year

1950
1970
1990
2010
0
100
200
300
400
DATABASE FEATURES

TOC

IC

DOC

Weighted by

drainage area

DATABASE FEATURES

Drainage Area
Discharge
0
20
40
60
80
100
IC
TOC
DOC
Percent included

in lateral flux database

0
10
20
30
40
50
IC
TOC
SUMMARY OF FINDINGS

Observed

Corrected for

drainage area

Corrected for

discharge

Lateral flux

Tg C yr
-
1

28

32

34

11

9

10

Total =

37

Total =

42

Total =

45

COMPARISON WITH OTHER STUDIES

0
5
10
15
20
25
30
This study
Mulholland 1982
Boyer et al. (In prep.)
Meybeck 1981
Schlesinger and Melack 1982
TOC Flux from Conterminous

U.S. (Tg C yr
-
1
)


-

Disaggregated by watershed area.



§

-

Disaggregated by biome type.

§

§

Inorganic carbon yield

(g C m
-
2

yr
-
1
)

Dissolved organic carbon yield

(
g C m
-
2

yr
-
1
)

> 7

4


7

2
-

4

< 2

No data

> 4

2


4

0.8


2

< 0.8

No data

REGIONAL PATTERNS

Carbon yield

g C m
-
2

yr
-
1

St Lawrence
Mississippi
Southwest
Northeast
Northwest
Eastern Gulf
Western Gulf
Southeast
Red
Colorado
0
2
4
6
8
Entire U.S.
IC

TOC

Calculating CO
2

flux


Stream surface area


CO
2

concentrations


Transfer velocity

Flux = stream surface area * ([CO
2
]
water
-

[CO
2
]
air
)*
k

High Resolution NHDPlus data


Inventory of streams


Stream order


Modeled discharge


Modeled slope


Modeled velocity

Stream area

Use modeled discharge to
calculate stream width.


Sum all stream segments
to obtain total stream area.

Stream area

ln discharge

n

= 1,026

Verification data from USGS
streamgaging stations

Raymond et al.
In Prep
.

Use modeled discharge to
calculate stream width.


Sum all stream segments
to obtain total stream area.

Stream area

ln discharge

n

= 1,026

Verification data from USGS
streamgaging stations

Raymond et al.
In Prep
.

Surface Area of
Streams: 40,560 km
2



~1/2 of Lake Superior
(US)

~Costa Rica


Calculated from
alkalinity, temperature,
and pH



4,200 sites



440,000 observations

CO
2

concentrations

CO
2

concentrations

Northern

K
600

(m d
-
1
) = 2841(SV) + 2.025


S= Slope ; V = Velocity

Gas transfer velocity

Raymond et al.

in prep

Gas transfer velocity

Very high
k
in
low
-
order
Western streams.

Total Flux:

96.7
±

32 Tg C yr
-
1

Total CO
2

flux from streams

Flux = stream surface area * ([CO
2
]
water
-

[CO
2
]
air
)*
k

Regional CO
2

yield

CO
2

yield

g C m
-
2

yr
-
1

Central
Gulf
Northern
Midwest
West
Southwest
0
10
20
Entire US
LATERAL VERSUS CO
2

FLUXES

Lateral flux

CO
2

CO
2

CO
2

CO
2

CO
2

CO
2

CO
2

flux


Diffuse


Dominated by low
-
order streams.



Lateral flux


Focused


Top 10 streams carry


75% of IC flux.


60% of TOC flux.

LATERAL VERSUS CO
2

FLUXES

CO
2
and lateral fluxes


Correlated with precipitation and
runoff.


Water discharge.


Carbon delivery from terrestrial
environment.


In
-
stream carbon
transformations.

Eastern Gulf

Northeast

Southeast

Southwest

Miss. R
.

Red R.

N. Pacific

W.Gulf

St.
Lawrence

Colorado

0
0.5
1
1.5
2
2.5
3
3.5
0
200
400
600
CO
2

r
2

= 0.86

TOC

r

= 0.72


g C m
-
2

yr
-
1


Runoff (mm yr
-
1
)

Total carbon flux in rivers

Total flux


(Tg C yr
-
1
)


IC

34

TOC

11

CO
2

97

Total

142

142 Tg C yr
-
1

Total carbon flux in rivers

Yield

(g C m
-
2

yr
-
1
)


IC


4.0

TOC


1.9

CO
2

14.9

Total 20.8

20.8 g C m
-
2

yr
-
1


Carbon fluxes in context

>142 Tg
C yr
-
1

Stream flow

45 Tg C
yr
-
1


97 Tg C yr
-
1

“CO
2

Flux”

“Lateral
Flux”

Sedimentation

?

Terrestrial and aquatic carbon flux

NEP

(g C m
-
2

yr
-
1
)


Grasslands

24
±

14

Zhang et
al. 2011

Total flux

Rivers

&
Streams

20.8

This study

Predictor of Basin Flux

Slope

r
2

Stream Surface Area %
1

0.72

0.78

Average Annual
Precipitation
2

23.1

0.86

% Forested Land
3

0.56

0.64


Precipitation controls the regional
differences in CO
2

flux



Short
-
term: flushing of soil CO
2




Long
-
term: geomorphology of the
density of stream networks
(stream surface area).


Regional CO
2

yield

LATERAL VERSUS CO
2

FLUXES

Lateral flux

CO
2

CO
2

CO
2

CO
2

CO
2

CO
2

CO
2

flux


Dominated by fluxes in small
-
order streams.


Strong correlation with precipitation.

Carbonate
Weathering

Soil
Respiration

Acidity:

Precipitation /
Mining

Allochthonous

DOM Respiration

Wetlands /
Riparian
Vegetation


The dominant source of pCO
2

will depend on scale


Headwater systems will show terrestrial respiration


Large river systems will incorporate internal and
external sources.


Contribution of Soil CO
2
to River Efflux


Soil Water pCO
2

20,000


30,000 ppm


Derived land surface run
-
off from USGS


Waterwatch


Wolcock et al.


in prep.


Total US discharge 1722 km
3


Amazon discharge ~ 5000 km
3




Region

Basin Area
thousand km
2

Predicted q
(m)

Predicted Max
Total Flux Tg C

Predicted CO
2

Evasion Flux Tg C

Max % of CO
2

efflux
exported laterally
(Soil)

Central

1020

0.36

6.6

23.3

28.1%

Northern

909

0.39

6.3

15.5

40.8%

Midwest

1965

0.06

2.2

15.3

14.4%

Gulf

1425

0.32

1.72

26.8

30.5%

West

1788

0.22

8.1

12.6

56.2%

Southwest

722

0.02

7.1

3.3

7.0%

Total US

7828

0.22

31.0

96.7

32.1%

Contribution of Soil CO
2

to River Efflux

Total carbon flux in rivers

Yield (g C m
-
2

yr
-
1
)

IC


3.9

TOC


2.8

CO
2

25.5

Total 32.2

Stream w =
e
(0.42(lnQ)+2.55)


Raymond et al. in prep

Stream Surface Area:

F(g C yr
-
1
) = ([CO
2
]
water
-

[CO
2
]
air
)* k *
Surface Area

1026 sites

REGIONAL PATTERNS

Northeast
Southeast
Eastern Gulf
Western Gulf
Mississippi
St Lawrence
Red
Colorado
Southwest
Northwest
0
2
4
6
8
All
Carbon yield

g C m
-
2

yr
-
1

REGIONAL PATTERNS

0
10
20
30
40
50
IC
POC
DOC
Observed

Corrected for

drainage area

Corrected for

discharge

Lateral flux

Tg C yr
-
1

28

32

7

34

3

8

3

6

3

Gulf
West
Central
North
Midwest
Southwest
0
10
20
Entire US
Regional CO
2

yield

CO
2

yield

g C m
-
2

yr
-
1

DATABASE FEATURES

'40
'50
'60
'70
'80
'90
'00
'10
'20
0
200
400
600
800
1000
Observations

per year

Willamette River

(at Salem, OR)

Illinois River

(at Peoria, IL)

(at Valley City, IL)

Schuylkill River

(at Philadelphia, PA)

Acknowledgements



Whitney Broussard (U of Louisiana, Lafayette)

Thor Smith (USGS Vermont / New Hampshire)

Valerie Kelly (USGS Oregon)

Evan Hornig (USGS Austin, TX)

Kate Halm (USGS Colorado)

Sources of data and methods of
analysis



Sources

Water quality
-


Clarke (1924) compilation of Dole & Stabler survey
(USGS).


U.S. Geological Survey (National Water Information
System).


Illinois State Water Survey.


Illinois Environmental Protection Agency.


Philadelphia Water Department.


Stream discharge



U.S. Geological Survey.




Methodological considerations



Nitrate analysis



Pre
-
1970

Colorimetric phenoldisulphonic acid method.

After 1970

Colorimetric cadmium reduction.


Organic nitrogen




“Albuminoid ammonia” method used on most samples before 1940.

Total N = Alb. NH
3

+ NO
3

+ NO
2

+ NH
3



Kjeldahl nitrogen used on all samples since 1975.

Total N = TKN + NO
3

+ NO
2


Sample handling


“Composite samples”


time
-
averaged.



Discrete samples


single time point.



Holding time in older samples.

Willamette River at Salem, OR

Drainage area

19,000 km
2


Human impacts in the watershed


Slaughterhouse and animal processing
facilities.


Urban wastewater and runoff.


Pulp and paper mill waste.


Logging / forestry.


Some agriculture in lower basin.


Noted historical water quality problems


Low oxygen concentration around
Salem (largely solved by 1972 through
discharge permit system).


Untreated urban wastewater
(successive rounds of treatment plant
updates 1930s through 1960s).


Depleted salmon population (ongoing,
improvements after 1970).

Illinois River

Drainage area

41,000 km
2

(upper)




69,000 km
2

(lower)


Human impacts in the watershed


Diversion of Chicago River (1890).


Urban runoff and wastewater.


Intense row crop agriculture.


Bottom land farms maintained through
levee & drainage districts.


Industrial effluent.


Noted historical water quality problems


Water
-
borne disease outbreaks in late
1800s.


Fisheries virtually destroyed by 1920s.


Low oxygen associated with
slaughterhouse wastes in Pekin/Peoria
area.


Schuylkill River

Drainage area

4,902 km
2



Human impacts in the watershed


First municipal water works in the nation
(1802).


Mine waste in upper watershed (1860s).


Heavy industry (1880s).


Moderate agriculture in lower watershed
outside of Philadelphia.


Intense urban development in lower
watershed.


Noted historical water quality problems


Numerous government reports between
1866 and 1946 recommended
abandoning the Schuylkill as a drinking
water source.


Coal mine silt pollution in upper basin.


Dramatic water quality improvements
resulted in the Schuylkill River being
designated as a Federal Scenic River
by 1978.