30
20
CHNL INFLW
CHNL OUTFLW
10
0
50
100
150
Time (h)
Figure 25. Inflow and outflow from a fictitious channel segment. This figure
shows the results of the implementation of the Muskingum channel routing
method in Object-GAWSER. CHNL_INFLW is the inflow to a channel seg-
ment. CHNL_OUTFLW is the outflow from the channel segment.
CHNL_INFLW peaks at 83 hours and CHNL_OUTFLW peaks at 91 hours.
The difference in the time and magnitude of the peaks shows that Object-GAW-
SER simulates the storage of water in a channel segment. Finally, the dip in
CHNL_OUTFLW at 60 hours represents the "Muskingum dip" referred to in
Chang et al. (1983).
Figure 25 describes the inflow and outflow from the channel segment. CHNL_OUTFLW is con-
stant until shortly before 60 hours where it begins to dip down while CHNL_INFLW begins to
increase. CHNL_OUTFLW begins to increase at 60 hours and peaks just after 90 hours.
CHNL_INFLW peaks higher and earlier than CHNL_OUTFLW thereby indicating channel storage.
The dip in CHNL_OUTFL illustrates one aspect of the Muskingum method in that outflow hydro-
graphs dip below the level of constant flow before beginning to increase again (Chang et al. 1983).
Because each sector represents a different area in a watershed, the sectors were linked to simulate
the movement of water from one area in a watershed to another area in a watershed. Sectors were
linked using connectors and ghosts. Connectors are described in Preliminary Description of STELLA
II Objects. Richmond (1994) describes ghosts as follows:
The Ghost is a replica of a level, flow, or converter. A replica is not a `copy,' as the term is
used in the Copy and Paste sense. A Copy has an independant identity because it possesses its
own underlying equation. By contrast, a replica is simply an image of the building block
from [which] it was ghosted.
The name, brief description, and units of the objects featured in this section are listed in Table 16.
There are two links between SNOMLT and GROFF1. The first link between SNOMLT and
GROFF1 simulates the flow of meltwater from the snowpack to impervious surfaces in the water-
shed. The first link was created by placing a ghost of LIQ_WTR_REL (the output from SNOMLT)
within GROFF1 and linking the ghost to RI_1 (the input to GROFF1) with a connector. The second
link was created to simulate rainfall on the impervious surfaces in a watershed that are not covered
by snow. The second link was created by placing a ghost of RAINs from SNOMLT in GROFF1 and
linking the ghost to RAINa. The value of RAINs is adjusted in RAINa. RAINa is linked to RI_1 with
a connector. RI_1 contains an equation that determines the amount of meltwater and rainfall that
will reach impervious surfaces in the watershed as a function of the percent of the watershed that is
not covered by snow (BARE).
38