1
MODELLING DISCHARGE AND NITROGEN FLUXES FROM
LARGE WATERSHEDS IN THE NORTHEASTERN UNITED
STATES D. P. Swaney1, R. W. Howarth1 , A. E..
Galford1, E.W. Boyer2, C.L. Goodale1 and R.M.
Marino1 1Cornell University, Ithaca, NY
USA 2University of California at Berkeley,
Berkeley, CA USA (email dps1_at_cornell.edu)
ReNuMa Hydrological Dynamics
Daily Temperature
Daily Precipitation
 ABSTRACT Much recent work has suggested that
nitrogen fluxes to the coastal zone depend on
interactions between input loading rates,
climate, and landscape processes. Direct
relationships between nitrogen inputs and
riverine fluxes have been shown with a relatively
simple nitrogen accounting methodology for 16
large watersheds in the northeastern US (Boyer et
al., 2002). More recently, statistical models
have shown that the average response of these
watersheds per unit nitrogen load is strongly
related to precipitation or hydrology (Howarth et
al., in press). To extend this approach to
examine the temporal response of these systems in
response to climate and land use change, we are
developing a simple Regional Nutrient Management
model (ReNuMa) based on watershed-scale water
balances and statistical relationships between N
loads and responses. For the 6-year period
examined so far, interannual discharge is well
predicted over the range of watersheds studied.
Nitrogen fluxes vary across watersheds in
response to anthropogenic sources, including
point sources, atmospheric deposition and
fertilizer use.
Snowpack
Evapotranspiration
Snowmelt
Urban
Streamflow
Shallow flow/runoff
SCS runoff equation for each cover type
Unsaturated zone
Annual DIN flux in 16 large Northeastern US
watersheds
Baseflow
Saturated zone
Across all watersheds, simulated annual DIN
fluxes also match observed values reasonably
well. Some individual watersheds exhibit
significant bias above or below observed fluxes,
but generally correlate with annual changes
variations reasonably well.
Human waste contributes N through sewers and
septic system effluent. N deposition traverses
lakes and wetlands without retention, but
exhibits a threshold landscape response in
forests. Agricultural N sources (fertilizer,
manure and fixation) also exhibit a threshold
response due to retention (ie landscape
denitrification, etc). The proportions of
in-river denitrification are based on estimates
in Van Breemen et al., 2002)
(1st order linear reservoir)
Annual streamflow in 16 large Northeastern US
watersheds
Simulated annual DIN fluxes improve as more
sources are added. Here, the scenarios show the
increase in several goodness-of-fit measures as
more terms are included in the model. Point
sources alone (5) underpredict the DIN load,
but still exhibit a significant R2 (ie a
linear relationship with observations). Adding a
fixed contribution from
Preliminary comparison of simulated vs observed
annual DIN fluxes for the 12 watersheds with
adequate observations of DIN ENS
Nash-Sutcliffe efficiency n6 years12
watersheds 72) ----------------------------------
-------------------------------------------- Run
Run description Bias R2 ENS -------------------
--------------------------------------------------
--------- 1 Baseline (all sources) 6.4 .89 .88
2 Pt sources agriculture -17.5 .86 .84
N dep (no forest) 3 Pt sources agriculture
-82 .90 .78 4 Pt sources agriculture
-142 .83 .57 without load response 5 Pt
sources only -225 .71 .13
16 watersheds in the Northeastern USA in which
Net Anthropogenic Nitrogen Inputs (NANI) have
been related to average riverine N fluxes
over the period 1988-93 (Boyer et al., 2002).
We are extending the analysis to
simulate seasonal and annual streamflow
and nitrogen fluxes using the ReNuMa model.
Examples of threshold responses in agricultural
and forest systems in the literature. Left
response of NO3 to fertilizer loads in
agricultural leachate (Billen Garnier, 2000)
Right response of NO3 to N deposition in
forest leachate (Aber, et al., 2003). We use a
similar parameterization to estimate landscape
response from these land cover types.
References The predecessor of ReNuMa is the
Generalized Watershed Loading Function Model
(GWLF) Haith, D. A., Shoemaker, L. L. 1987.
Generalized watershed loading functions for
stream flow nutrients. Water Resources Bulletin
23(3)471-478. A spreadsheet-based version of
the model can be found on the web at
http//cfe.cornell.edu/biogeo/USGSWRI.htm.
Other references
The 2928 weather stations in the National Climate
Data Center network for Northeastern states were
identified (http//www.ncdc.noaa.gov/oa/ncdc.html)
. To select candidate stations for each
watershed, Thiessen polygons for the network were
generated using ArcView 3.2. Stations with
polygons intersecting a watershed and with gt95
complete records (daily temperature and
precipitation) were averaged to obtain
representative weather data for each watershed.
Missing temperature data for each station were
replaced with averages of the records preceding
and following the missing interval missing
precipitation values were replaced with zero.
agricultural lands greatly improves the
agreement, and adding a load-dependent contributio
n improves it further. Direct N deposition and
the response of forests also improves the
agreement with observed fluxes across watersheds.
Aber, J. D., C. L. Goodale, S. V. Ollinger, M.-L.
Smith, A. H. Magill, M. E. Martin, R. A. Hallett,
and J. L. Stoddard. 2003. Is Nitrogen Deposition
Altering the Nitrogen Status of Northeastern
Forests? Bioscience 54(4)375-389. Billen, G.
and J. Garnier. 2000. Nitrogen transfers through
the Seine drainage network a budget based on the
application of the Riverstrahler model.
Hydrobiologia. 410 139150. Boyer, E.W., C. L.
Goodale, N. A. Jaworski and R. W. Howarth. 2002.
Anthropogenic nitrogen sources and relationships
to riverine nitrogen export in the northeastern
U.S.A. Biogeochemistry 57/58137-169.
Howarth, R.W., D.P. Swaney, E.W. Boyer, R.M.
Marino, N. Jaworski and C.L. Goodale. The
influence of climate on average nitrogen export
from large watersheds in the Northeastern United
States. Accepted for publication in
Biogeochemistry. Van Breemen, N., E.W. Boyer,
C.L. Goodale, N.A. Jaworski, K. Paustian, S.P.
Seitzinger, K. Lajtha, B. Mayer, D. Van Dam, R.W.
Howarth, K.J. Nadelhoffer, M. Eve, and G. Billen.
2002. Where did all the nitrogen go? Fate of
nitrogen inputs to large watersheds in the
northeastern U.S.A. Biogeochemistry 57/58
267293.
Ongoing work and future directions

• Our work to date has focused on developing
  parameterizations of biogeochemical responses
  related to landuse/landcover
• Atmospheric deposition
• Landscape level N retention of agricultural N
  sources (fertilizer, manure, N-fixation)
• Nitrogen retention and losses (DIN) from forests
• In-stream and landscape denitrification
• We are currently working on refining the model
  parameters for individual watersheds to reduce
  bias at this scale, as well as developing
  alternative parameterizations of processes which
  incorporate results from smaller-scale models.
  Additional processes to be considered will
  include
• Phosphorus losses from P-saturated soils
• Soil erosion and sediment transport

Minor adjustments to evaporative cover factor
for the Androscoggin and Susquehanna rivers were
the only changes made to baseline parameter
values for the watersheds. While individual
watersheds may exhibit some bias above or below
observations (ie USGS annual streamflows
http//waterdata.usgs.gov/nwis/ ), agreement over
all years and across all watersheds was generally
good.
Acknowledgements This work has been supported by
an EPA STAR grant, Developing regional-scale
stressor models for managing eutrophication in
coastal marine ecosystems, including interactions
of nutrients, sediments, land-use change, and
climate variability and change, EPA Grant Number
R830882, R.W. Howarth, P.I.