(1) Department of Physical Geography, Faculty of Geosciences, Utrecht University, Utrecht, The Netherlands

(2) Department of Earth Sciences – Geochemistry, Faculty of Geosciences, Utrecht University, Utrecht, The Netherlands

Global land-ocean linkage: direct inputs of water and associated nutrients

to coastal zones via submarine groundwater discharge (SGD)

Hans Dürr(1), R. van Beek(1), C.P. Slomp(2), H. Middelkoop(1), M. Bierkens(1)

h.durr@geo.uu.nl

20th SWIM20th SWIM24. June 200824. June 2008

Effects of nutrient enrichment (eutrophication) of coastal waters

red tide(harmful algal bloom)

foam

more algalblooms

Rationale• Higher nutrient inputs (N, P) from rivers to coastal waters +

estuaries, inverse trend (decrease) due to damming for Si

• Global nutrient fluxes at the land-ocean interface: Terrestrial controls, impact on coastal waters and response to global change. Spatially explicit, multiple nutrient forms (dissolved / particulate, inorganic / organic)

• Box modelling (Slomp & van Cappellen 2004) has shown potential impact on coastal zone of nutrient SGD at global scale

• Typically: Groundwater N and P flux = total SGD flux x nutrient concentration, but decisive for net balance: fresh GW flux x nutrient concentration ( local scale effects of subterranean estuary)

• Submarine Groundwater Discharge (SGD) of nutrients: spatially explicit not yet estimated at global scale, comparison to river inputs (water discharge globally ~5-10% of river discharge)?

Proximal coastal ocean

burial

recycling

river input

groundwaterinput

Phytoplankton (mostly N-limited)

export to continentalshelf

N:P ~ 14:1

N:P mostly>>> 16

eolianinput N

denitrification

Redfield N:P = 16:1

• Fresh SGD: mostly N, less P (elevated velocity seeps in shallow, oxic systems: P removal, conservative behaviour of NO3), could drive systems to P-lim.• Only in cases when both GW and sediment are anoxic and upon saltwater intrusion: N/P ratio of SGD < Redfield• Deep GW supply mostly minor & not contaminated

Snow cover

QBf

Canopy

Store 2

Store 1

Store 3

PRECEpot

Eact

QDR

QSfP

P

T

River ice

TwTA

RS

Snow/Rain

QChannel

PRECEpot

QChannel

TA

RS

Snow/Rain

LH(Epot)

Q

Courtesy R. van Beek

Model structure

0,5° grid

Each cell: - Conceptual model partly

based on the HBV model

- Climate variables are from the ERA40 data set

- Transfer / Routing of runoff to the drainage network

For baseflow here: - Not fully calibrated yet- Energy balance + lakes will be

considered in later stages- Monthly values annual means

Concept of groundwater flow• Kraaijenhof-Van de Leur (1958)

– Boussinesq – Dupuit assumptions to calculate the reservoir coefficient kr

where: k is the hydraulic conductivity (L·t-1 e.g., m/day), D is the aquifer thickness (L, 50m assumed), f is the drainable porosity (obtained from lithology + some Holdridge climate data) (dimensionless, 0-1) and B is the aquifer width (L).

Using various global data sets

2

2

4 fBkDkr

Courtesy R. van Beek

Residence time of Groundwater

Courtesy R. van Beek

Long residence times: - Highly weathered soils - Tropics

Result:

Model performance (no full calibration yet): - Good in most temperate and tropical regions- Weaker in Arctic + semi-arid regions: no evaporation from surface waters yet, snowmelt driven by mean monthly temp.

Next step – full calibration on low flow conditions

Local groundwater discharge

Courtesy R. van Beek

High in: Some tropical regions, esp. SE-Asia / Indonesia; W Canada, Florida; Ex.: Patos Lagoon

Low: Arctic, but % contribution to total flow is high

Next steps, including nutrients

• Coastal water divide where baseflow = SGD??• Human GW extraction? • Salt-water intrusion sites (low-lying areas)?!

• Range of GW nutrient concentrations for DIN ~NO3• Test nutrient attenuation scenarios, e.g., based on half-life

approaches • Coupling to information from landuse and population

Coastal ribbon definition for SGD: distance to water divide at coast

Problem :

what is a stream ???

every catchment basin that has a cumulated upstream area > xx km2

testing possible

0,5° grid cell

subgrid variability

last gauging station

Courtesy Rens van Beek

Water extraction: - country based GW abstraction from IGRAC / TNO - linked to total water use and population numbers from Vörösmarty et al.

Groundwaterabstraction

USGS Summary data

Range ofGW nutrient

concentrationsin different

landusesettings

(NO3)

Agricultural Urban Misc. Undevel. land

Data from IGRAC / TNO

High Groundwater Nitrate sites

Potential hotspots with potentially elevated N input

Caution: - Very local phenomena not detected - Total Groundwater flow in coastal cells, not SGD - GW abstraction not considered - Saltwater intrusion not considered- Effect of residence time in GW on nutrient concentrations

Some study sites for SGD including nutrients

DIN-NEWS empirical model: Dumont et al. 2005

Rivers

Groundwater

Similarities in regions, local differences

Conclusions: first steps towards• Spatially explicit estimates of SGD at global scale, using

baseflow estimates from new global hydrological model• Potential hot spots of freshwater nutrient SGD

– SE Asia (esp. Indonesia) + Central America: hot spots for both river and GW fluxes (high baseflow, high runoff and high levels of human activity) – exact locations and time scales may be different

• Field studies needed in high ‘risk’ areas, e.g. (sub-)tropical areas of Africa, S America and SE Asia – some studies now underway

• Examples have been shown to include amendments such as GW abstraction, GW quality data, landuse + population data

As source of ‘new’ nutrients, especially N (less

P), freshwater SGD is potentially important for coastal nutrient cycling

(N-limitation of most coastal waters) at global scale (strong response to

human impact)

Thank you !

Questions ?Hand-outs available, ask anytime or email:

h.durr@geo.uu.nl

Salt-water intrusion: CaHCO3 freshwater is replaced by NaCl and SO4

2- rich sea water

Ca2+Ca2+

Na+

Ca2+

Ca2+ Na+Na+

-Na2+ replaces Ca2+, NH4+

-Cl- or SO42- replaces PO4

- Increased degradation of organic matter (Nyvang, 2003)- Increased sulfate reduction & reduction of Fe-oxides by S2- NH4 and PO4 release to the groundwater

Na+