model coefficients using each calibration dataset (i.e., observed TN loads in few of
days) and models with the lowest Akaike information criterion (AIC) values were
selected for load estimations.
The reliability of DC-HSPF is at the heart of the hydrological simulation for daily
river discharge. First, in addition to natural conditions, human activities (e.g., landuse and land-cover changes, reservoirs, transboundary water transfers, human water
uses, wastewater discharges) in the watershed were completely taken into consideration in the DC-HSPF. Second, the effect of lake water withdrawals was taken into
account and assumed to be entirely used for cropland irrigation in the direct runoff
areas, rather than the river basins that are located far from the lake (Fig. 12.1). Third,
the parameters of DC-HSPF were calibrated by observations of 16 rivers from 2009
to 2011 and further validated by routine daily observations for the 2 largest rivers
over the period 1999–2008 as well as by a detailed investigation of the Yunnan
Hydrological Bureau (YHB). Last, an automatic method for parameter estimation
(PEST) was applied for calibrating model parameters (Wu et al. 2018). The detailed
model framework of the DC-HSPF is provided in Wu et al. (2018). The predictive
performances of the DC-HSPF were measured using three quantitative statistics:
Nash-Sutcliffe efficiency (NSE), percent bias (BIAS), and ratio of the root mean
square error to the standard deviation of measured data (RSR).
12.3 Fluxes of N Depositions and Riverine Inputs
12.3.1 Dry N Deposition
The total N flux of dry deposition over the Lake Dianchi was 42.1 Æ 10.3 mg km
À2
s
À1 in 2010–2011 (1σ as the standard deviation of fluxes occurring in 5 sites;
Fig. 12.2a). For gaseous and particulate N, the deposition flux is 37.1 Æ 10.5 mg
km
À2 s
À1 and 5.0 Æ 0.7 mg km
À2 s
À1
, respectively. Ninety-nine percent of the dry
deposition was NH 3 (32.8 Æ 10.1 mg km
À2 s
À1
, 77.9%), HNO 2 /HNO 3 (4.0 Æ 0.6 mg
km
À2 s
À1
, 9.5%), particulate ON (2.9 Æ 0.6 mg km
À2 s
À1
, 6.8%), and particulate
NO 3
À -N (2.0 Æ 0.1 mg km
À2 s
À1
, 4.7%). Corresponding concentrations and V d of
gases and TSP at five sites can be found in Figs. 12.3 and 12.4. Dry deposition N flux
showed strong seasonality regardless of three missing samples in early April–June
(Fig. 12.2b). Fluxes from May to August were 1.7–13.3 times higher than the
remaining periods. From May to August, dry deposition of NH 3 was the major source
(>86.0%) of total dry deposition N flux, with biweekly coefficient of variation [CV] of
115%. The NO 2 flux was remained low in early spring and summer (0.2 Æ 0.1 mg
km
À2 s
À1
) but high in autumn and early winter (0.4 Æ 0.2 mg km
À2 s
À1
). The
remaining N fluxes were negligible with a seasonal variability with CV of ~27%.
A Nationwide Nitrogen Deposition Monitoring Network (NNDMN) containing
43 monitoring sites was established in China to measure gaseous NH 3 , NO 2 , and
HNO 3 and particulate NH 4
+ and NO 3
À in air from 2010 to 2014 (Xu et al. 2015).
Two of 43 sites over vegetation were located close to Lake Dianchi and were
12 Impacts of Nitrogen Deposition on China’s Lake Ecosystems. . .
271
days) and models with the lowest Akaike information criterion (AIC) values were
selected for load estimations.
The reliability of DC-HSPF is at the heart of the hydrological simulation for daily
river discharge. First, in addition to natural conditions, human activities (e.g., landuse and land-cover changes, reservoirs, transboundary water transfers, human water
uses, wastewater discharges) in the watershed were completely taken into consideration in the DC-HSPF. Second, the effect of lake water withdrawals was taken into
account and assumed to be entirely used for cropland irrigation in the direct runoff
areas, rather than the river basins that are located far from the lake (Fig. 12.1). Third,
the parameters of DC-HSPF were calibrated by observations of 16 rivers from 2009
to 2011 and further validated by routine daily observations for the 2 largest rivers
over the period 1999–2008 as well as by a detailed investigation of the Yunnan
Hydrological Bureau (YHB). Last, an automatic method for parameter estimation
(PEST) was applied for calibrating model parameters (Wu et al. 2018). The detailed
model framework of the DC-HSPF is provided in Wu et al. (2018). The predictive
performances of the DC-HSPF were measured using three quantitative statistics:
Nash-Sutcliffe efficiency (NSE), percent bias (BIAS), and ratio of the root mean
square error to the standard deviation of measured data (RSR).
12.3 Fluxes of N Depositions and Riverine Inputs
12.3.1 Dry N Deposition
The total N flux of dry deposition over the Lake Dianchi was 42.1 Æ 10.3 mg km
À2
s
À1 in 2010–2011 (1σ as the standard deviation of fluxes occurring in 5 sites;
Fig. 12.2a). For gaseous and particulate N, the deposition flux is 37.1 Æ 10.5 mg
km
À2 s
À1 and 5.0 Æ 0.7 mg km
À2 s
À1
, respectively. Ninety-nine percent of the dry
deposition was NH 3 (32.8 Æ 10.1 mg km
À2 s
À1
, 77.9%), HNO 2 /HNO 3 (4.0 Æ 0.6 mg
km
À2 s
À1
, 9.5%), particulate ON (2.9 Æ 0.6 mg km
À2 s
À1
, 6.8%), and particulate
NO 3
À -N (2.0 Æ 0.1 mg km
À2 s
À1
, 4.7%). Corresponding concentrations and V d of
gases and TSP at five sites can be found in Figs. 12.3 and 12.4. Dry deposition N flux
showed strong seasonality regardless of three missing samples in early April–June
(Fig. 12.2b). Fluxes from May to August were 1.7–13.3 times higher than the
remaining periods. From May to August, dry deposition of NH 3 was the major source
(>86.0%) of total dry deposition N flux, with biweekly coefficient of variation [CV] of
115%. The NO 2 flux was remained low in early spring and summer (0.2 Æ 0.1 mg
km
À2 s
À1
) but high in autumn and early winter (0.4 Æ 0.2 mg km
À2 s
À1
). The
remaining N fluxes were negligible with a seasonal variability with CV of ~27%.
A Nationwide Nitrogen Deposition Monitoring Network (NNDMN) containing
43 monitoring sites was established in China to measure gaseous NH 3 , NO 2 , and
HNO 3 and particulate NH 4
+ and NO 3
À in air from 2010 to 2014 (Xu et al. 2015).
Two of 43 sites over vegetation were located close to Lake Dianchi and were
12 Impacts of Nitrogen Deposition on China’s Lake Ecosystems. . .
271