Page 361 - Soil and water contamination, 2nd edition

348 Soil and Water Contamination

a 6 0.20 b 4
6955
Phosphorus (mg P l -1 ) 5 4 3 0.10 Flow (m 3 s -1 ) Phosphorus (mg P l -1 ) 2
3
0.15

1 2 0.05 1
0 0.00 0
22 January 2002 23 January 2002 24 January 2002 25 January 2002 0 0.05 0.10 0.15 0.20
Flow (m 3 s -1 )
Flow Total P Reactive P (< 0.45 µm)

Figure 18.8 Response of different forms of P to storm events in the Den Brook catchment, Devon, UK, in January
2002. Adapted from Haygarth et al. (2004).

The hysteresis pattern can change from storm to storm and across seasons (Walling and
Webb, 1986). For example, when the store of readily soluble substances in soil has become
exhausted over time due to progressive leaching or uptake by vegetation during the growing
season, the concentration response may decrease during the successive storm events. On the
other hand, mineralisation of litter at the end of the vegetative period may have the opposite
effect. Rising groundwater levels during winter and spring can also influence hysteresis
patterns, as the contribution of groundwater to flow increases. Furthermore, variations in
hysteresis between storm events can also be caused by different patterns of rainfall across
the catchment . Consequently, although hysteresis patterns can be generalised, the Q–C
relationship in a river may not follow the same pattern between storms.
18.3.3 Release of old water

Storm events not only cause increases in inputs from overland flow , but may also mobilise
substantial amounts of ‘old’ or pre-event water, which may even dominate the increase in
discharge. Eshleman et al. (1992) performed a chemical hydrograph separation based on

chloride concentrations as a means of quantifying the absolute and relative contributions
of ‘new’ rainwater and ‘old’ groundwater to the storm hydrograph. Figure 18.9a shows
-
a hydrograph and chemograph of Cl during a 5.2 cm rainfall event in the Reedy Creek
catchment in Virginia, USA, during spring 1991. The rainstorm caused substantial
-
-
Cl dilution in the creek, since Cl concentrations are much lower in precipitation than
-
in baseflow . Provided that the Cl concentrations in rainwater and groundwater do not
-
vary during the storm or vary spatially across the catchment, measurements of the Cl
concentration in the two components and in the mixture over time can be used to back-
calculate the contributions from ‘old’ and ‘new’ water. Figure 18.9b shows that ‘old’ water
is almost totally responsible for the increase in discharge in this catchment during the storm
event. Though negative, the hysteresis loop is clockwise (Figure 18.7c), which implies that
the main contribution of event water is during the falling limb of the hydrograph. This
contradicts the general assumption that the rising limb of the hydrograph is dominated by
channel precipitation and quick overland flow, as explained in the previous paragraph. This
contradicts the general assumption that the rising limb of the hydrograph is dominated by
channel precipitation and quick overland flow, as explained in Section 18.3.1. In contrast to
2+
+
-
the Cl concentration, the concentrations of reactive chemical species (such as Ca or H )
in old water may be highly variable. The combination of rapid mobilisation of old water

10/1/2013 6:47:12 PM
Soil and Water.indd 360 10/1/2013 6:47:12 PM
Soil and Water.indd 360