estimate of ΔQ. Therefore, further efforts at making long-term measurements of dry
deposition using relaxed eddy accumulation systems (Toya et al. 2007) are needed.
In addition, the contribution of indirect N deposition on watersheds needs to be
determined.
Lastly, groundwater discharge, as an additional N input to the lake, was also not
included because no measurements were available. We estimated it using a lake
water balance approach for the period of April 2010–March 2011:
G in ¼ ET lake þ ΔS þ Q out þ C out À P lake À Q in
ð12:9Þ
where G in was the groundwater discharge in lake, m
3
Ámonth
À1 ; Q in was the monthly
total streamflow from all rivers, m
3
Ámonth
À1 ; and ΔS denoted changes in water
storage in lake within a month (+, increase; À, decrease), m
3
Ámonth
À1 , which was
determined based on observed stage-storage curves ( f(D)) for Lake Dianchi, where
D was lake water depth, m; P lake was precipitation over Lake Dianchi, which was
determined as area-weighted mean of five meteorological stations around Lake
Dianchi, m
3
Ámonth
À1 ; ET lake was lake evaporation, m
3
Ámonth
À1 ; Q out was monthly
outflow from the lake at Haikou Station (Fig. 12.1), m
3
Ámonth
À1 ; C out was humaninduced water consumption of withdrawal from the Lake Dianchi (e.g., irrigation) to
watershed, m
3
Ámonth
À1 ; Q in , D, P lake , and Q out were obtained from local observations; and C out was converted from electricity consumption by pumping for agricultural irrigation and industrial use, where electricity consumption was obtained
through a field survey conducted by our group. To reduce the uncertainty due to
model structure, ET lake was estimated as average values of the results calculated by
the Penman-Monteith equation (Min 2001), mass transfer equation (Shuttleworth
1993; Singh and Xu 1997), and the pan evaporation equation (Linacre 1993),
where most of coefficients were determined based on local empirical equations
(Min 2001). Eventually, groundwater discharge was calculated based on lake
water balance for the period of April 2010–March 2011. Figure 12.15 indicates
that groundwater discharge (3.6 Â 10
6 m
3 year
À1 ) is much lower than streamflow
(380.9 Â 10
6 m
3 year
À1 ). Although uncertainties exist in the estimates of ET lake
and C out , the contribution from groundwater discharge may be negligible for total
N inputs.
12.6 Main Findings and Outlook
Atmospheric deposition of N proved to be one of the important sources of N input
over Lake Dianchi, especially during initial proliferation and periods of maximum
Microcystis bloom formation. Similar results could be found in Lake Taihu — the
third largest freshwater lake in China. Ti et al. (2018a) investigated wet and dry
deposition of N, including gaseous NH 3 , NO 2 , and HNO 3 and particulate NH 4
+ -N
and NO 3
À -N in both the atmosphere and precipitation during a 2-year period at three
sites in the Lake Taihu region in southeast China. The total N deposition in this
12 Impacts of Nitrogen Deposition on China’s Lake Ecosystems. . .
289
deposition using relaxed eddy accumulation systems (Toya et al. 2007) are needed.
In addition, the contribution of indirect N deposition on watersheds needs to be
determined.
Lastly, groundwater discharge, as an additional N input to the lake, was also not
included because no measurements were available. We estimated it using a lake
water balance approach for the period of April 2010–March 2011:
G in ¼ ET lake þ ΔS þ Q out þ C out À P lake À Q in
ð12:9Þ
where G in was the groundwater discharge in lake, m
3
Ámonth
À1 ; Q in was the monthly
total streamflow from all rivers, m
3
Ámonth
À1 ; and ΔS denoted changes in water
storage in lake within a month (+, increase; À, decrease), m
3
Ámonth
À1 , which was
determined based on observed stage-storage curves ( f(D)) for Lake Dianchi, where
D was lake water depth, m; P lake was precipitation over Lake Dianchi, which was
determined as area-weighted mean of five meteorological stations around Lake
Dianchi, m
3
Ámonth
À1 ; ET lake was lake evaporation, m
3
Ámonth
À1 ; Q out was monthly
outflow from the lake at Haikou Station (Fig. 12.1), m
3
Ámonth
À1 ; C out was humaninduced water consumption of withdrawal from the Lake Dianchi (e.g., irrigation) to
watershed, m
3
Ámonth
À1 ; Q in , D, P lake , and Q out were obtained from local observations; and C out was converted from electricity consumption by pumping for agricultural irrigation and industrial use, where electricity consumption was obtained
through a field survey conducted by our group. To reduce the uncertainty due to
model structure, ET lake was estimated as average values of the results calculated by
the Penman-Monteith equation (Min 2001), mass transfer equation (Shuttleworth
1993; Singh and Xu 1997), and the pan evaporation equation (Linacre 1993),
where most of coefficients were determined based on local empirical equations
(Min 2001). Eventually, groundwater discharge was calculated based on lake
water balance for the period of April 2010–March 2011. Figure 12.15 indicates
that groundwater discharge (3.6 Â 10
6 m
3 year
À1 ) is much lower than streamflow
(380.9 Â 10
6 m
3 year
À1 ). Although uncertainties exist in the estimates of ET lake
and C out , the contribution from groundwater discharge may be negligible for total
N inputs.
12.6 Main Findings and Outlook
Atmospheric deposition of N proved to be one of the important sources of N input
over Lake Dianchi, especially during initial proliferation and periods of maximum
Microcystis bloom formation. Similar results could be found in Lake Taihu — the
third largest freshwater lake in China. Ti et al. (2018a) investigated wet and dry
deposition of N, including gaseous NH 3 , NO 2 , and HNO 3 and particulate NH 4
+ -N
and NO 3
À -N in both the atmosphere and precipitation during a 2-year period at three
sites in the Lake Taihu region in southeast China. The total N deposition in this
12 Impacts of Nitrogen Deposition on China’s Lake Ecosystems. . .
289