Escolar Documentos
Profissional Documentos
Cultura Documentos
1 of 8
https://ceprofs.tamu.edu/folivera/TxAgGIS/Spring2002/verwest/verwest.htm
LITERATURE REVIEW
Original attempts to link geographic information systesm (GIS) to hydrologic models began in the mid 1970s (Nunes et al.,
1998). Not until the early 1990s when GIS increased its functionality did hydrology really take advantage of this new
technology. Three different approaches were made to integrate GIS: modeling completely with current GIS tools, bridging data
from GIS to another program, and embedding code into GIS to create new software (Kopp, 1996). GIS expanded the
possibilities of hydrologic models since it can handle such a large amount of data.
GIS is capable of handling large amounts of data for processing and eliminates repetitively (Goonetilleke and Jenkins, 1999).
Because of the large amounts of topographical and geological data stored in maps, GIS lends itself to the perfect tool since it
can handle the quantity of spatial data (Chang et al., 2000). By using GIS to create the data for inputs, estimations improve
since a larger amount of data can be used to calculate the parameters (Schumann et al., 2000) and a more objective view of the
data is taken (Elgy et al., 1993). Some considerations need to be taken into account when using GIS for analysis such as the
problem to be analyzed, the required output, and the data available. This can influence what type of GIS software is used.
Also, whether the model is tightly or loosely coupled is another consideration (Goonetilleke and Jenkins, 1999).
Most current hydrologic models that take advantage of GIS link to other programs to perform the hydrologic analysis. The way
that hydrologic analysis is done using GIS is often referred to as either loosely or tightly coupled with any variation in between.
Loosely coupled data is when GIS is used primarily for data organization, and the data is then transferred to a hydrologic model
for analysis; tightly coupled data is when all analysis is done in one program, either the hydrologic model or GIS, with a link to
the other program to perform limited operations. Tightly coupled analysis, on the other hand, is limited because of the complex
interface, and loosely coupled analysis is more common. Loose coupling is commonly done by performing initial analysis in
GIS, converting this information into a useable form for the hydrologic model, performing the hydrologic analysis, and finally
transferring the information back to GIS for display (Hellwegger and Maidment, 1999).
9/7/2012 7:39 PM
2 of 8
https://ceprofs.tamu.edu/folivera/TxAgGIS/Spring2002/verwest/verwest.htm
One example of a loosely coupled hydrologic model is defined in Hellweger and Maidment. The three software components
used were GIS software, software to convert GIS data into hydrologic data, and hydrologic modeling software. The GIS
software is basic commercially available software that is used to gather and analyze the data needed in the hydrologic model.
In order to communicate to the hydrologic model, new software was developed to communicate with both GIS and the
hydrologic modeling software. The software used in this paper was CWR-PREPRO to collect vector properties and prepare an
ASCII file needed for input into the hydrologic model. The final software used was the Hydrologic Modeling System available
through the Hydrologic Engineering Center, which performs the hydrologic analysis. With the type of hydrologic software, and
therefore the type of hydrologic model, defined, a method was developed to automate the process (Hellwegger and Maidment,
1999). The analysis cannot be reasonably applied to areas of flat terrain, so a new hydrologic model and system of processing
data must be developed.
METHODOLOGY
Most hydrologic models are designed for average or steep terrain. Flat terrain has unique problems that make this analysis
unsuitable. A couple of these problems are that water does not always travel downhill and channels can be bi-directional.
Because of this, using the digital elevation model (DEM) to delineate the watershed is not practical and hydrologic models that
can handle only dentritic networks cannot be used.
In flat terrain, water does not necessarily travel downhill because a slight change in elevation, such as a curb or a plant, can
change the direction of steepest slope. Also, friction plays a larger role. Water will take the path of least resistance, so a
slightly lower slope with less friction may be chosen over a higher slope with greater friction. In order to handle the way that
water flows in flat terrain, an existing channel network was used. Water was assumed to flow to the nearest channel in the
network. Using Euclidean distance, the contributing area for each channel was determined. The output of finding the
contributing areas was a grid as seen in Figure 1.
9/7/2012 7:39 PM
3 of 8
https://ceprofs.tamu.edu/folivera/TxAgGIS/Spring2002/verwest/verwest.htm
9/7/2012 7:39 PM
4 of 8
https://ceprofs.tamu.edu/folivera/TxAgGIS/Spring2002/verwest/verwest.htm
Area (ft2)
5000
Maximum Distance
(ft)
832
1600
254
5400
917
3200
533
400
70
Since both the subbasins and the longest flow path were determined using a grid, the size of the grid for the study area must be
determined. Two components were considered to establish the grid size: the computer processing time and the detail of study
necessary. For the computer processing time, the amount of time it took to create the subbasin and distance grid were both
measured. For the detail of study, the minimum channel length, and the number of channels with no subbasin area after
analysis were considered. These parameters were gathered for a range of resolutions, and then the grid size for the final
analysis was found. With the subbasins delineated and some hydrologic parameters known, the other properties for final
analysis needed to be calculated.
Most parameters such as velocities will be determined based on general characteristics of the entire watershed. One parameter
that is available is the curve numbers for the area. This can be obtained from a curve number grid for the entire United States.
The study area is then clipped from the larger grid. Once the water is routed into the channels, some consideration needs to be
made about how the water will flow in the channels.
Besides travel time in the reaches of the channel, the direction the water flows in needs to be considered. In flat terrain,
channels can be bi-directional, meaning that depending on the amount of rain in a certain location, water can flow either
direction in a channel as seen in Figure 4.
9/7/2012 7:39 PM
5 of 8
https://ceprofs.tamu.edu/folivera/TxAgGIS/Spring2002/verwest/verwest.htm
Resolution (ft)
Rows
1000
500
320
641
00:04
00:10
9/7/2012 7:39 PM
6 of 8
https://ceprofs.tamu.edu/folivera/TxAgGIS/Spring2002/verwest/verwest.htm
400
801
885
00:12
105
300
1,068
1,180
00:16
59
200
1,602
1,770
00:27
29
100
3,203
3,541
01:19
10
50
6,407
7,081
04:52
25
12,814
14,162
21:00
From Table 2, the processing time can be seen to vary from 4 seconds to 21 minutes. The number of channels that have no
subbasin area vary from 699 to 0. The processing time was balanced against the number of channels that were not assigned an
area. The shortest channel in the study had a length of 71 feet, and only 4 channels in the network are less than 100 feet in
length. Based on this information, a resolution of 100 feet was chosen for the rest of the analysis as seen in Figure 6.
9/7/2012 7:39 PM
7 of 8
https://ceprofs.tamu.edu/folivera/TxAgGIS/Spring2002/verwest/verwest.htm
7,178,310
Curve
Number
71
Maximum Distance
(ft)
2,900
906
13,354,875
70
2,800
918
919
17,169,214
5,014,550
73
70
2,400
2,100
920
922
47,720,120
6,919,442
75
78
5,886
2,400
923
924
3,232,782
9,790,568
86
82
1,910
2,400
GIS ID
Area (ft2)
901
Curve numbers were then found for the subbasins. Using area-weighted average, the curve numbers were calculated for the
subbasins from the polygons. The resulting curve numbers can be seen in Table 3. The next steps would be to create the final
hydrologic model, run the model, and compare the outputs of the model to the gage stations.
CONCLUSIONS
Flat terrain has unique characteristics, which make typical hydrologic analysis impossible. By developing a new way of
9/7/2012 7:39 PM
8 of 8
https://ceprofs.tamu.edu/folivera/TxAgGIS/Spring2002/verwest/verwest.htm
delineating the watershed, the way in which water flows to channels is redefined. Also, the unique flow velocities of overland
and channel flow is taken into account to find the longest flow path of the subbasin. This process is not compatible with
previous hydrologic models, so a whole new program for final analysis must be created. This takes time to create and test the
new model. Once the model is ready, flat terrain will have a more suitable way of being represented.
REFERENCES
Chang, Tsang-Jung, M.H. Hsu, W.H. Teng, and C.J. Huang. A GIS assisted distributed watershed model for simulating flooding
and inundation. Journal of the American Water Resources Association. Vol. 36, n. 5, Oct 2000, p. 975-988.
Elgy, J., C. Maksimovic, and D. Prodaovic. Matching standard GIS packages with urban storm drainage simulation software.
Application of Geographic Information Systems in Hydrology and Water Resources Management: Proceeding of the
HydroGIS93 Conference in Vienna, Austria. International Association of Hydrological Sciences (IAHS) Publication
No. 211, April 1993, p. 151-160.
Goonetilleke, A., and G.A. Jenkins. The role of geographical information systems in urban hydrological modeling. Journal of
the Chartered Institution of Water and Environmental Management. Vol. 13, n. 2, Jun 1999, p. 200-206.
Hellwegger, F.L., and D.R. Maidment. Definition and connection of hydrologic elements using geographic data. Journal of
Hydrologic Engineering. Vol. 2, n. 1, Jan 1999, p, 10-18.
Kopp, Stephen M. Linking GIS and hydrological models: where we have been, where we are going? Application of
Geographic Information Systems in Hydrology and Water Resources Management: Proceeding of the HydroGIS96
Conference in Vienna, Austria. International Association of Hydrological Sciences (IAHS) Publication No. 235, April
1996, p. 133-139.
Nunes Correia, Francisco, F. Castro Rego, M. da Graca Saraiva, and I. Ramos. Coupling GIS with hydrologic and hydraulic
flood modeling. Water Resources Management. Vol. 12, n. 3, Jun 1998, p. 229-249.
Schumann, A.H., R. Funke, and G.A. Schultz. Application of a geographic information system for conceptual rainfall-runoff
modeling. Journal of Hydrology. Vol. 240, n. 1-2, Dec 2000, p. 45-61.
9/7/2012 7:39 PM