Page 206 - Biosystems Engineering

184 Cha pte r F i v e

The interface was designed for use by the U.S.D.A. Natural Resource
Conservation Service (NRCS) State Office personnel as a screening
tool in formulating water quality project plans for impaired water-
sheds and for evaluating the impacts of conservation and manage-
ment practice implementation on NPS pollution. Invariably, all these
researchers concluded that interfacing GIS with NPS models pro-
vided a tool for efficient management of temporally and spatially
referenced data and allowed flexible interactive model runs, analy-
ses, and visualization.

5.4.1 Approaches for Interfacing GIS
with Watershed Models
Interfacing GIS with distributed parameter hydrologic models is
playing an increasingly important role in the management of water
quality and water resources as well as in designing, calibrating, mod-
ifying, and comparing these models. There are several strategies and
approaches to interface spatially distributed environmental models
and GIS. They range from simple pre- and postprocessor linkages
through shared data files to building models as complex analytical
functions into fully functional GIS, or embedding the required GIS
functionality into spatially distributed models (Fedra 1995). The
interfacing of models, especially watershed models, is not a novel
concept anymore and has been tried numerous times since the 1990s.
However, as pointed out by Vckovski et al. (1999) and Martin et al.
(2005), the lack of consistent data protocols and the lack of interest by
commercial software providers in creating universal data transfer
standards has stifled the development of standardized frameworks
for terminology, data exchange formats and interfacing procedures.
The following terminology describe interfacing efforts: “couple,”
“link,” “combine,” “interface,” “model within,” and “integrate”
(Martin et al. 2005).
As described by Martin et al. (2005), linking typically means that
data generated by GIS is used as input by the models, and output is
transferred back to GIS for display and spatial analysis. This is typi-
cally done through ASCII and binary data file transfers. This approach
limits users to take full advantage of the spatial analysis capabilities
of a GIS because it is mainly used for pre- and postprocessing for
input and output data. In combining, information is passed between
the model and GIS via memory-resident data models rather than
external files (Liao and Tim 1997). Martin et al. (2005) stated that
“This approach improves computational performance and interactiv-
ity between the two software systems, translating into a more sophis-
ticated modeling environment.” Although data transfer between
models is automated and hidden from the user, the interactivity
remains somewhat limited.