Geoscience Reference
In-Depth Information
Erosion fromAgricultural Management Systems (CREAMS), the Groundwater Loading Effects of
Agricultural Management Systems (GLEAMS) and the Erosion Productivity Impact Calculator
(EPIC). EPIC later evolved into the Environmental Impact Policy Climate model and is a direct
descendant of the Simulator for Water Resources in Rural Basins (SWRRB) model which aimed
to predict the responses of runoff and sediments to management impacts of ungauged rural
basins across the USA. The latest version of the model is SWAT2009.
As their names suggest, however, these models were aimed at improving agricultural man-
agement, of which runoff prediction was only one component - albeit an important one, in
that runoff acts as a driver for many of the transport processes considered. In the USA, the
model has been used to simulate all of the catchment areas in the conterminous states for
more than a decade in support of national policy development. It has also been widely used in
different parts of the world, because it provides a flexible tool for rainfall-runoff modelling and
agricultural management that, very importantly, has been made freely available (see Appendix
A). At the time of writing, some 600 articles have been published that describe using SWAT.
A publication list can be found at the SWAT web site. Gassman
et al.
(2007) give a tabular
review of many different applications of SWAT.
B6.2.2 Definition of Hydrological Response Units in SWAT
Each catchment in a SWAT application can be subdivided into subcatchments linked by the
channel network. In the USA, the subcatchments are based on the eight-digit hydrologic cat-
aloguing units (HCU) that subdivide the conterminous USA into 2149 catchment units.
Each subcatchment can then be subdivided into a number of hydrological response units
(HRUs) of homogeneous land use, management and soil characteristics. The HRUs are treated
as fractions of the local subcatchment and need not be differentiated in space. Runoff predicted
from each HRU is routed directly to the channel network. The hydrological components of
each HRU include interception; partitioning of precipitation, snowmelt and irrigation water
between surface runoff and infiltration; redistribution of water in the soil profile; evapotran-
spiration from the root zone; lateral subsurface flow in the soil profile; and return flow from
shallow groundwater.
Fast runoff is predicted by using the Soil Conservation Service Curve Number method
(SCS-CN) at daily or sub-hourly time intervals or the Green-Ampt infiltration equation us-
ing sub-hourly rainfall rates. Redistribution in the soil profile uses a storage-flux method be-
tween specified soil layers, with the possibility of preferential flow in soils with secondary
structure (Arnold
et al.
, 2005). Three methods of estimating evapotranspiration are also avail-
able, depending on available weather data: the Penman-Monteith equation, the Priestly-Taylor
equation and the Hargreaves equation.
Recharge below the soil profile is partitioned between shallow and deep groundwater sys-
tems. The shallow groundwater can produce return flow back to the surface and can also lose
water to the atmosphere through deep rooting plants. Water that reaches the deep groundwa-
ter is assumed not to contribute to streamflow but is treated as a loss to the catchment water
balance. A routine is included for accounting for the effects of field drainage on soil water and
runoff (Green
et al.
, 2006). Runoff from the different mechanisms included in the hydrology
component of the model are summed and routed through the channel network using either a
variable-rate storage method or the Muskingum method.
The availability of the SWAT code has led to a number of variants on the model with modified
components and interfaces with different GIS systems and databases (including the ESRI Arc
packages) and Windows GUIs (see Gassman
et al.
, 2007) in defining HRUs. Santra
et al.
(2011)
have suggested a fuzzy set approach to defining HRUs that they test using the SWAT model.
Recent developments include a simplified HRU representation in which the partitioning of
fluxes is based on the water balance for use in data scarce regions (Easton
et al.
, 2010; White
et al.
, 2011).
