 |
| Titel |
Temporal and spatial variations in groundwater quality resulting from policy-induced reductions in nitrate leaching to the Rabis Creek aquifer, Denmark |
| VerfasserIn |
Søren Jessen, Peter Engesgaard, Lærke Thorling, Sascha Müller, Jari Leskelä, Dieke Postma |
| Konferenz |
EGU General Assembly 2016
|
| Medientyp |
Artikel
|
| Sprache |
en
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 18 (2016) |
| Datensatznummer |
250136479
|
| Publikation (Nr.) |
 EGU/EGU2016-17523.pdf |
|
|
|
|
|
| Zusammenfassung |
| Twenty-five years of annual groundwater quality monitoring data from the sandy unconfined
Rabis Creek aquifer were used to assess the effects of political actions aimed to reduce nitrate
leaching to the aquifer. Data were collected from eight multilevel samplers along a ∼3 km
transect, which follows the general direction of groundwater flow. Each multilevel sampler
comprises 20 screens placed with a 1 m vertical distance from near the water table
downwards. The transect covers areas of livestock, plantation & heath, and agriculture. The
history of nitrate leaching to the aquifer was assessed using data from screens close to the
water table of multilevel samplers placed within agricultural areas. According to these data,
nitrate concentrations of infiltrating ‘agricultural’ water peaked at 2-3 mM (120-180
mg/L) in the year 1989, and then gradually decreased and stabilized at 0.25-1.0 mM
(15-60 mg/L) from year 2000. Local farmers declare having used the maximum
fertilization rate allowed during the period. The timing of the observed decrease therefore
suggests a direct link to the political action plans implemented in the same period.
Parallel to the development in nitrate leaching, although with a transport time lag,
the average concentration of nitrate in the oxic zone of the aquifer was roughly
halved between 2000 and 2013. As a response to political initiatives of the late
1980’ies, part of the area covering the aquifer was changed from agriculture to
non-fertilized grass for livestock; the data shows that this effectively remediated the aquifer
underneath in less than 20 years, to become nitrate-free and attain background
sulfate levels. The oxidized and pyritic reduced zone of the aquifer is separated by
a <1 m redoxcline. Denitrification by the pyrite releases sulfate that is retained
down-gradient as a tracer for historical nitrate loading to the aquifer. Thus sulfate
concentrations currently increase in the multilevel samplers positioned furthest
down-gradient along the transect. During the 25 year monitoring period the redoxcline has
moved by one to a few decimeters, as controlled by the aquifer sediment’s pyrite
content. Further, the data indicate that no zero-valent sulfur is precipitated during
pyrite oxidation in the aquifer, while most of the pyritic iron is precipitated. Nickel
(Ni2+) is released at the redoxcline resulting in concentrations more than twice
the 20 μg/L Danish drinking water limit. The data clearly indicate that this Ni2+
contamination can be ascribed to the agricultural nitrate loading and would not
occur under natural conditions. A 2D reactive transport model was constructed
(PHAST 3) to simulate the temporal and spatial development in nitrate and sulfate
concentrations in the aquifer while taking into account effects of dispersion. The model
predictions indicate that sulfate concentrations, despite dispersive mixing, is still
increasing along down-gradient stretches of the aquifer, where flow paths surface from
the deeper up-gradient part of the aquifer, to eventually discharge into the Rabis
Creek. |
|
|
|
|
|
|