Agriculture Reference
In-Depth Information
scales maintaining systematic processes,
but may not include detailed geospatial
structure and coupled hydrodynamic pro-
cesses of the larger catchment and river
basin. Other models are available that repre-
sent some variables of interest based on stat-
istical inference, plot or hybrid approaches
(Arnold
et al
., 1998, SWAT model; Stöckle
et al
., 2003, CropSyst model; Schwarz
et al
.,
2011, Sparrow model). The intention is not
to propose a new model, but rather to inte-
grate process-based plot-scale models with
a catchment-scale model that fills a gap for
efficient modelling of lateral transport to
first- and second-order streams while also
allowing simulation scales of mesoscale
catchments.
A strategy for merging scales and con-
cepts intrinsic to plot-, landscape- and
catchment-scale biogeochemical research is
proposed. An integrated hydrodynamic ap-
proach to numerical modelling can resolve
the geospatial characteristics of water and
carbon and nitrogen cycles over entire
catchments. The intention is that the meso-
scale model will have application for scales
that range from
10
-2
to 10
5
km
2
. The current
status of modelling N and C dynamics can
be described in terms of two types:
the average response in the HRU. In this
approach, the lateral mass transfer uses a
one-way coupling strategy to fluid flow.
This approach preserves the
1-
D continuum
characteristics including vertical macropo-
re effects, but is limited to empirical lateral
mass transport, which limits the use of dis-
tributed validation data sets (e.g. ground-
water and baseflow data).
Ideally, the next generation of simulation
models should expand on the experience
gained in the development of these models
and incorporate a modern scientific and soft-
ware infrastructure. It should be biophysical-
ly based, with fully distributed hydrology,
have a common object-oriented structure to
simulate nutrient flow and vegetation pro-
cesses, and transport across land units, and
the ability to simulate the variations in
flows generated by management. The mod-
elling infrastructure outlined above follows
a 'bottom-up', process-based approach to
simulating landscape-level effects. Processes
occurring in land segments, and landscape-
level processes governed by the movement
of water in watersheds, need to be repre-
sented to simulate water quality at any point
in the landscape.
Technically, it is progress in this direc-
tion that will allow a seamless integration
between stakeholder demands and science
delivery in the form of model outputs. In
this context, a new
2-
D continuum approach
for lateral surface runoff and groundwater
flow with embedded and fully coupled
1-
D
soil profile processes is proposed. This ap-
proach is still not a fully
3-
D continuum
strategy, since these are limited to relatively
small-scale computational domains and our
practical goal here is a catchment-scale model.
It recognizes the extreme computational re-
quirements of the
3-
D hydrobiogeochemical
continuum by a suitable dimension reduc-
tion, but still makes progress at the catchment
scale (
10
-2
to 10
5
km
2
).
As an example of this strategy, one
could use a catchment hydrodynamic model,
the Penn State Integrated Hydrologic Model
(PIHM;
http://www.pihm.psu.edu
) that is
currently being implemented at Critical
Zone Observatory (CZO) sites in the USA
1.
The soil profile model approach where
detailed vertical distributions of water, en-
ergy and solutes are explicitly modelled
(
1-
D continuum) while plant growth/decay,
landscape features, soil are represented as a
discrete domain defined for some character-
istic of the landscape or an average over some
area. At this scale, lateral spatial gradients are
not explicitly resolved. A primitive version
of this type uses vertical averaging to repre-
sent the soil in terms of the bulk water, en-
ergy and solute dynamics. The soil profile
can include vertical macropore-matrix ef-
fects (Gerke and van Genuchten, 1993b) and
complete kinetic descriptions of the coupled
water and biogeochemical reactions.
2.
A semi-distributed model approach com-
bines the concept of hydrologic response
units (HRUs) and the
1-
D continuum ap-
proach from (1) to divide the catchment into
characteristic landscape units that allow
mass transfer to adjacent sub-areas based on
Search WWH ::
Custom Search