Determination of the drainage basin characteristics using vector GIS

H. Apaydin1, F. Ozturk1, H. Merdun2 and N.M. Aziz3 1 Department of Agricultural Engineering, Faculty of Agriculture, Ankara University, Diskapi, Ankara 06110, Turkey. E-mail: [email protected] 2 Department of Agricultural Engineering, Faculty of Agriculture, Kahramanmaras Sutcu Imam University, Kahramanmaras 46060, Turkey 3 Department of Civil Engineering, Clemson University, Clemson, SC 29634, USA Received 9 August 2004; accepted in revised form 5 October 2005 Abstract Detailed geomorphologic characteristics need to be compiled for performing hydrologic modeling of a basin. Basin form and hydrologic characteristics are to be related so the basin form must also be represented by quantitative descriptors. The typical morphologic characteristics used in hydrological analyses are basin area, perimeter, mainstream length, total stream length, contour length, basin shape (form factor, circularity ratio, compactness ratio, basin elongation), slope, drainage density, relief (maximum relief, relief ratio, relative relief), effective basin width, and median elevation. The objective of this study is to propose an algorithm to automatically calculate basin characteristics using vector GIS. The results produced by the algorithm were compared to the manual method and the two methods were found statistically similar. Nordic Hydrology Vol 37 No 2 pp 129 –142 q IWA Publishing 2006 Determination of the drainage basin characteristics using vector GIS Keywords Basin morphologic characteristics; basin shape; basin slope; drainage density; hydrologic modeling Introduction Basins are areas divided by natural hydrological boundaries and used to manage water resources and develop solutions to environmental problems. These areas include assemblages of natural resources that rely on the type and quantity of water present within the basin (Reimold 1998). The main characteristics of a drainage basin are that it functions as an open system, its function changes over short periods of time, there are numerous relationships between the morphologic elements and the processes occurring in the drainage basin, and these relationships are complex and multivariate in character. A complete understanding of the drainage basin characteristics requires measurements, analyses and relationships of the processes operating in the basin. Technological developments help to satisfy these purposes such as field examination, laboratory models, or knowledge of physical or statistical laws by analogy (Gregory and Walling 1973; Eash 1993b; Dehn et al. 2001). Whichever of these three approaches is used, measurements must be obtained of the drainage basin form, which includes the overall character of the basin as well as of the component parts such as the stream network and the stream channel. The possible interrelationships between hydrology and morphometry are seemingly infinite and the parameters are related in a complicated manner so that equations are unable to explain all the variability. Time of hydrograph rise and lag time will be shorter and peak discharge rate may be higher in the basins with the highest relief ratio. Similarly, the orientation of a pear-shaped basin with respect to the outlet or contrasted bifurcation ratios will doi: 10.2166/nh.2006.004 Downloaded from http://iwaponline.com/hr/article-pdf/37/2/129/364736/129.pdf by guest 129 H. Apaydin et al. 130 determine whether the hydrograph will peak early or late as seen in Figure 1. It is possible to increase examples showing the relationship between individual morphologic characteristics and the hydrograph. Figure 2 is a schematic model showing the expected influence of variations in basin characteristics on the hydrograph based on the generalizations from a large number of studies as summarized later. In each case, the influence of a specified characteristic variable is displayed assuming that all other morphometric, geologic, and climatic variables remain constant (Gregory and Walling 1973; Ritter et al. 1995). The shape and character of a stream hydrograph should be affected greatly by the manner in which a basin collects and routes water through its network. Flow is significantly related to many components of basin and network morphometry, but they are very hard to express as a single value (Ritter et al. 1995). Carlston (1963) demonstrated a very close relationship between drainage density and mean annual flood in 15 small basins in the USA. It was also found that the rate of base flow was inversely related to drainage density in large basins. Morisawa (1967) establish a relationship (Q ¼ a L b ) between mean annual discharge (Q) and length of longest stream (L) using 96 basins in 6 different locations of the eastern USA. Glymph and Holtan (1969) illustrated different types of relationship which can be obtained. Figure 1 The effect of basin relief and shape to hydrograph (Gregory and Walling 1973) Downloaded from http://iwaponline.com/hr/article-pdf/37/2/129/364736/129.pdf by guest H. Apaydin et al. Discharge Hydrograph 1 Hydrograph 2 Time Basin characteristics Hydrograph type 1 Hydrograph type 2 Basin area Small Large Drainage density High Low Basin magnitude High Low Relief High Low Ruggedness number High Low Basin shape Equidimensional Elongate Soils Thin Thick Vegetation Sparse Dense Figure 2 Hydrograph and generalized responses to the drainage basin characteristics (Ritter et al. 1995) Patton and Baker (1976) demonstrated predictive relationships between basin magnitude, channel frequency, drainage density, ruggedness number, relief ratio, and peak flow in several regions of the USA. Their regression equations predicted peak flow up to the prediction accuracy of R 2 ¼ 0.92. They also found that basins with the high flash flow potential had greater ruggedness numbers than the low potential basins. Costa (1987) investigated the morphometry of basins associated with the largest historic floods in the United States and found that basins with peaked flood hydrographs generally contained significant areas of exposed bedrock, occurred in semiarid to arid climates, were short, and had high relief. More recent studies use a set of morphological variables. Sefton and Howarth (1998) attempted to determine the relationships between dynamic response characteristics and physical descriptors of catchments (i.e. mean elevation, mean slope, area, stream frequency, etc.) in England and Wales. A simple land use scenario demonstrated an application of the methodology in which variation in physical descriptors of catchments may be used to assess the impacts of environmental change. Berger and Entekhabi (2001) determined the degree to which the basin’s morphological features and regional climate can explain or predict hydrologic response. The relationships between physical characteristics (median slope, relief ratio, drainage density, infiltration capacity, wetness ratio and saturated zone efficiency index) and the hydrologic properties of basins were investigated and these characteristics estimated a runoff ratio with an R 2 of 0.90 in 10 basins in the USA. Sankarasubramanian and Vogel (2002) developed a relationship between runoff ratio and potential evapotranspiration, relative infiltration capacity, average slope, and drainage density based on observed runoff ratios at 1305 basins across the USA with an R 2 of 0.71. The influence of watershed shape was found to be significant in several studies conducted by the USGS where they statistically related climatic and watershed properties to stream discharge data. The 100 year recurrence interval discharge, Q100 (m3/s), is related to the area (A), the elevation E (m), and the basin form factor (BSFF) as Q100 ¼ 0:471A 0:715 E 0:827 BSFF 0:472 (Ward and Trimble 2004). In this view automated extraction of Downloaded from http://iwaponline.com/hr/article-pdf/37/2/129/364736/129.pdf by guest 131 H. Apaydin et al. 132 basin characteristics are crucial in the development of reliable hydrologic models for a basin. These basin characteristics must be supplemented by measurements of the processes operating as input to, as processes within, and as output from, the basin. Recent improvements in computer hardware and software technology have enabled software developers to automate this process. Traditionally, basin characteristics such as basin area, drainage density, relief, and basin shape have been derived manually from maps, aerial photographs, and field surveys. These techniques are tedious, costly, time consuming, and subject to considerable operator variance. Recently, Geographical Information Systems (GIS) technology has been increasingly employed to assist hydrologists with the task of model parameterization. GIS is a computer based tool that allows input, storage, analysis, and display of the spatial data. GIS is being used by resource managers, earth scientists, urban planners, civil and environmental engineers, among others for inventory analysis, estimation, planning, and modeling (FAO 1988; Aronoff 1989; Bhaskar et al. 1992). The GIS links land data to topographic data and other information concerning processes and properties related to geographic location. Once the database is constructed, correlations between different pieces of information can be examined through statistical and geostatistical analyses and computer generated maps. It is clear that the potential value of GIS applications to hydrologic modeling and assessment justifies the continued study of this technology (Aronoff 1989; DeVantier and Feldman 1993). Harvey and Eash (1996) developed Basinsoft to quantify 27 morphometric characteristics (i.e. drainage area, basin length, perimeter, slope, relief, shape factor, channel length, stream order, etc.) of a drainage basin. The program was written in Arc Macro Language and requires three coverages (a drainage divide, streams, and contours) and a lattice elevation model. Statistical comparison tests indicated that the Basinsoft quantifications were not different from the manual topographic map measurements for 9 of 10 basin characteristics tested. Doan (2000) introduced HEC-GeoHMS, which has a capability of determining stream length, slope, basin area, and slope between endpoints, etc., developed in The Hydrologic Engineering Center. Ries (2002) introduced STREAMSTATS developed by the USGS. STREAMSTATS works through a web browser to provide various types of streamflow statistics for data-collection stations and for ungauged sites. It determines the watershed boundaries and measures physical and climatic characteristics of the watersheds for the ungauged sites using GIS, and then inserts the characteristics (like area, relief and slope) into previously determined regression equations to estimate the streamflow statistics. Numerous methods of describing drainage basins have been proposed; some of these apply to the whole basin, while the others apply to a particular characteristic, such as relief or soil. Any single index is not normally adequate because it attempts to simplify complex reality and express, often in two dimensions or in a single index, what the three-dimensional reality and time magnitude as well as time significance are (Gregory and Walling 1973; Linsley et al. 1988). Even though GIS techniques are superior to the manual methods in the quantitative determination of basin characteristics, a number of problems or difficulties remain. Perhaps one of the most important ones is that the general purpose GIS typically lacks the capability to easily derive a number of hydrologically significant basin variables from topographic data (Spence et al. 1995; Lacroix et al. 2002). Over the past decades, researchers have demonstrated the viability of techniques for automatically deriving a wide variety of topographic and topologic basin information directly from digital data. The automatic derivation of basin data is faster, less costly, and more reproducible than traditional, manual Downloaded from http://iwaponline.com/hr/article-pdf/37/2/129/364736/129.pdf by guest H. Apaydin et al. techniques and provides data in a digital form that can be readily imported and analyzed in GIS (Jenson and Domingue 1988; Lacroix et al. 2002). As the applications of GIS to many different scientific fields increased, the use of the GIS in hydrologic studies in basin scale also substantially increased. In most of these studies, since raster data collection through remote sensing devices is easier and in some cases a vector database is not available, raster systems are commonly used. This study using vector systems is believed to contribute to the determination of basin morphologic characteristics. Therefore, the objective of this paper is to present an algorithm (Baschar) for calculating basin characteristics such as basin area, perimeter, main stream length, total stream length, contour length, basin shape (form factor, circularity ratio, compactness ratio, basin elongation), slope, drainage density, relief (maximum relief, relief ratio, relative relief), effective basin width, and median elevation, automatically using only vector data. Verification of the algorithm is conducted by comparing the results of the algorithm with the basin characteristics obtained manually. Materials and methods The basins in the study area are the Ankara –Guvenc basin, the Canakkale –Egridere basin, the Kutahya– Kocacesme basin, the Sanliurfa– Kizlar basin and the Tokat –Akdogan basin located in Turkey. The basins have different characteristics such as area, perimeter, and slope. The study areas are presented in Figure 3. Unit hydrographs of basins are given in Figure 4. The Kocacesme basin has a karstic soil structure and very low flow reached the basin outlet; thereby it was impossible to draw hydrographs of this basin. These basins were selected from the National Soil and Water Resources Research program executed by General Directorate of Rural Services (KHGM) (Karas 2001; Aykanli et al. 2002; Kaya and Helaloglu 2003; Oguz and Balcin 2003; Tekeli and Babayigit 2003). Topographic attributes of drainage basins may be visualized in different scales such as the basin as a whole, total channel system or network, individual parts of channels or reaches, and channel cross section. For each of these four categories, indices of area, length, shape, and relief are required. Some of these values are absolute: area, length, perimeter, relief, but are subject to operational definitions. Some are obtained by combining two absolute measurements as in density or measures of slope and the others require a definite method for Figure 3 The locations of the selected basins in Turkey Downloaded from http://iwaponline.com/hr/article-pdf/37/2/129/364736/129.pdf by guest 133 H. Apaydin et al. Figure 4 Unit hydrographs of the basins their computation as in basin shape, or drainage network (Gregory and Walling 1973). The algorithm developed in this study to determine these characteristics requires both ArcInfo and ArcView 3.0. The algorithm calculates the basin characteristics by using four vector coverages. These coverages are basin boundary, main stream, branches, and contours. Table 1 and Figure 5 present the necessary ArcInfo coverages for the algorithm. ArcInfo process ArcInfo is required for digitizing and editing. Four different data coverages were used in the simulation using a digitizer (Figure 5). The names of the coverages must be the same as those in Table 1. In the contour coverage, both contours and basin boundaries must be digitized. After the digitization process, two different topologies must be built in the contour coverage, first the polygon and second the line topology. Consequently, elevation values of contours should be stored at contour line coverage. Once the coverages were completely built, there was no more need for ArcInfo. ArcView process The procedures described below were implemented in Avenue scripts in ArcView 3.0. The scripts read information in the format of ArcInfo and calculate the characteristics of the basin. While the algorithm is running, project name, project directory, map scale, contour interval, and elevation of the highest and lowest point should been entered. Finally, the basin characteristics are displayed as well as written to the hard disk of the computer (Figure 6). Estimation of characteristics Area, perimeter, main stream length, total stream length, and contour length. Drainage basin area is in many respects the easiest basin characteristic to relate to drainage basin processes Table 1 Required ArcInfo coverages Coverage name 134 BASINB MAINSTR BRANCH CONTOUR Downloaded from http://iwaponline.com/hr/article-pdf/37/2/129/364736/129.pdf by guest Coverage description Coverage type Basin boundary Main stream Branches of stream Contours and basin boundary Polygon Line Line Line and polygon H. Apaydin et al. Figure 5 Required ArcInfo coverages for the algorithm but as it is in turn correlated with other characteristics its significance may not always be easy to interpret. In the manual method, basin area was measured by planimeter. Similarly, main stream, total stream, and contour lengths were measured by map meter. In this study, the four coverages were digitized from the 1/25 000 scale topographic map. The basin area, and the individual stream and contour lengths, were automatically calculated by the ArcInfo as the basins were digitized, while the total stream and contour lengths were calculated by the software developed after the digitizing finished (ESRI 1994, 1995, 2001). These values were corrected by the algorithm using the scaling factor if necessary. Basin shape. Shape is one of the most important topographic properties being measured with accuracy. The shape, drainage network, and channel system of a basin can all be influenced significantly by other drainage basin characteristics, such as rock type. These characteristics can affect the basin processes, particularly in that they may determine the potential efficiency of the basin, the network or the channel. The shape of a basin also affects the streamflow hydrograph and peak flow rates (Gregory and Walling 1973; Linsley et al. 1988). Several methods express the shape of basin. In this study the form factor, circularity ratio, compactness ratio, and basin elongation were used as measures of basin shape. Form factor. The shape of a drainage basin, as it is projected upon the horizontal datum plane of a map, may conceivably affect stream discharge characteristics. Horton (1941) regarded the pear shape as one proof that drainage basins were formed largely by sheet-erosion processes acting upon an initially inclined surface (Chow 1964). Horton (1932) suggested the dimensionless form factor (BSFF) as an index of shape: BSFF ¼ A L2 ð1Þ where A is the basin area and L is the length of the basin measured from outlet to divide near the head of the longest stream along a straight line. As the stream channel may be very sinuous, the length parallel to the main drainage line has been used to give a measure of basin length (Chow 1964; Gregory and Walling 1973; Linsley et al. 1988). For example for a circle Downloaded from http://iwaponline.com/hr/article-pdf/37/2/129/364736/129.pdf by guest 135 H. Apaydin et al. Figure 6 The sample output screens of the algorithm BSFF ¼ 0.79, for a square with the outlet at the midpoint of one side BSFF ¼ 1, and for a square with the outlet at one corner BSFF ¼ 0.5 (Linsley et al. 1988). Circularity ratio. Dimensionless circularity ratio (BSCR) is defined as the ratio of basin area (A) to the area of a circle (ACircle) having the same perimeter as the basin (Chow 1964; Gregory and Walling 1973): BSCR ¼ A A ¼ ACircle pr 2 ð2Þ The perimeter of the circle must be equal to the perimeter of the basin (P): P 2p ð3Þ A A 4pA BSCR ¼
2 ¼
P 2  ¼ 2 : P P p 4p 2 p 2p ð4Þ P ¼ 2pr ) r ¼ Thus Compactness ratio. Dimensionless compactness ratio (BSCM) is defined as the ratio of the perimeter of the basin (P) to the perimeter of a circle (PCircle) of equal area (Eash 1993a, b; Harvey and Eash 1996): BSCM ¼ P P ¼ : PCircle 2pr The area of the circle must be equal to the area of the basin: rffiffiffiffi A 2 : A ¼ pr ) r ¼ p ð5Þ ð6Þ Thus BSCM ¼ 136 P P P
A0:5 ¼
2 0:5 ¼ pffiffiffiffiffiffiffi : p A 2 pA 2p p 2 p As the compactness ratio decreases, the basin shape approaches a circle. Downloaded from http://iwaponline.com/hr/article-pdf/37/2/129/364736/129.pdf by guest ð7Þ H. Apaydin et al. Basin elongation. Dimensionless basin elongation (BSBE) is defined as the ratio of the diameter of a circle of the same area as the basin to the maximum basin length (L) (Chow 1964): 2r BSBE ¼ ð8Þ L The diameter of the circle must be equal to the area of the basin: rffiffiffiffi A A ¼ pr 2 ) r ¼ : ð9Þ p Thus qffiffiffi 2 pA : ð10Þ BSBE ¼ L This ratio ranges between 0.6 and 1.0 over a wide variety of climatic and geologic types. Values near to 1.0 are typical of regions of low relief, whereas values in the range 0.6 –0.8 are generally associated with strong relief and steep ground slopes (Chow 1964). Slope. The slope of a basin is defined as the division of the total fall between the highest and the lowest points to the basin length. Because of variation in land surface slope in the usual basin, a method of defining an average or index value is required. The slope of the ground surface is an important factor in the overland-flow process and hence a parameter of hydrologic interest, especially in small basins where the overland-flow process may be a dominant factor in determining hydrograph shape (Linsley et al. 1988). In this study, two different methods of slope determination were used. The first method is defined by Williams and Berndt (1972):   P LC h 12 ðLC 1 þ LCn Þ þ n21 i i¼2 SWB ¼ £ 100 ð11Þ A where SWB is the slope of basin in per cent, h is the contour interval (km), LC is the length of contour (km), and A is the basin area (km2) (Williams and Berndt 1972). The second method is defined by Wisler and Brater (1959): P h ni¼1 LCi SW ¼ £ 100 ð12Þ A where SW is the slope of the basin in per cent, h is the contour interval (km), LC is the length of contour (km), and A is the basin area (km2) (Wisler and Brater 1959; Black 1991). Area-elevation data. When one or more of the factors of interest in a hydrologic study vary with elevation, it is useful to know how the basin area is distributed with elevation. An area – elevation (or hypsometric) curve can be constructed by measuring the area between contours on a topographic map and plotting the cumulative area above (or below) a given elevation versus that elevation (Linsley et al. 1988). Drainage density. The total length of streams within a basin area divided by drainage area defines the drainage density, the average length of streams per unit area. Dimensionally, this ratio reduces to the inverse of length, L 21 (Chow 1964; Linsley et al. 1988; Black 1991): DD ¼ SLT A Downloaded from http://iwaponline.com/hr/article-pdf/37/2/129/364736/129.pdf by guest ð13Þ 137 H. Apaydin et al. where DD is the drainage density (km/km2), SLT is the total stream length (km) and A is the basin area (km2). Drainage density may be considered as the intensity of channels in an area, indicating the closeness of channels. If geometrical similarity exists between two drainage systems, their drainage densities are related to the same ratio as the inverse of the linear scale ratio. Thus, broadly considered, drainage density is simply one of several linear measures by which the scale of features of the topography can be compared (Chow 1964). A high drainage density reflects a highly dissected basin, which should respond relatively rapidly to a rainfall input, while a low drainage density reflects a poorly drained basin with slow hydrologic responses. This index has been widely adopted for its ease of comprehension, its simplicity, and its utility (Gregory and Walling 1973; Linsley et al. 1988). Relief. Relief, as a drainage basin characteristic, has been expressed in several ways and undoubtedly exercises an influence over runoff and sediment production in the basin. The influence derives from the fact that higher relief or steeper slopes potentially provide more available energy than do more subdued basins. Parameters of the stream flow hydrograph will vary according to basin relief and basin shape. The time of hydrograph rise and the lag time will be shorter, and peak discharge rate may be higher in basins with the highest relief ratio (Chow 1964). Maximum relief. Maximum relief, RM, is the elevation difference between the highest and lowest points (Chow 1964). Relief ratio. Dimensionless relief ratio (RR) is the ratio of maximum basin relief (RM) to horizontal distance along the longest dimension of the basin parallel to the principal drainage line (SLM) (Chow 1964): RR ¼ RM : SLM ð14Þ The possibility of a high correlation between relief ratio and hydrologic characteristics of a basin is suggested by Schumm (1954) who found that sediment loss per area is closely correlated with relief ratio (Chow 1964). Relative relief. Relative relief (RRE) in per cent is the ratio of maximum basin relief (km) to basin perimeter (km) (Eash 1993a, b; Harvey and Eash 1996): RRE ¼ RM £ 100: P ð15Þ Effective basin width. Effective basin width (EBW) (km) is the ratio of the area of the basin (km2) to basin length (km) (Eash 1993a, b; Harvey and Eash 1996): EBW ¼ A : L ð16Þ Basin length is measured along the main-channel, flood-plain valley from basin outlet to basin divide (Eash 1993b). 138 Median elevation. Median elevation (ME) can be found using a hypsometric curve. The median elevation of a basin is that elevation where half of the basin elevation data is higher Downloaded from http://iwaponline.com/hr/article-pdf/37/2/129/364736/129.pdf by guest and the other half is lower. Unlike average elevation, median elevation is more representative of the basin elevation because it provides more information on the elevation distribution within a basin (Linsley et al. 1988; Mohamoud 2004). Results and discussion Downloaded from http://iwaponline.com/hr/article-pdf/37/2/129/364736/129.pdf by guest H. Apaydin et al. Basin characteristics were estimated both manually and automatically by computer in five different basins. The results for both the manual and automated methods are tabulated in Table 2. The relationship between area and runoff volume indicates that, as area increases, runoff also increases as stated by Ward and Trimble (2004). The Kizlar and Guvenc basins are the two largest basins in terms of area among the five basins (Table 2) and the hydrographs of these two basins have the largest peak values (Figure 4). The Kizlar and Guvenc basins have similar values of peak runoff (4338 l/s and 5766 l/s, respectively). The Akdogan basin has the smallest area and peak value. The Egridere and Akdogan basins produce similar values (1745 l/s and 1565 l/s, respectively). Form factor, circularity ratio, compactness ratio, and basin elongation describe basin shape. Effective basin width may also be included into this group. The Kizlar and Guvenc basins are similar in shape (Figure 3). They also have the first two rank degrees in form factor, basin elongation and effective basin width among the five basin shape characteristics (Table 2). When basin shape is equidimensional as in the Kizlar and Guvenc basins (similar to Figure 1(F)), expected hydrograph type is 1 (in Figure 2) and the hydrograph itself is as seen in Figure 4. The other three basins (Akdogan, Egridere, and Kocacesme) have elongated shapes (Figure 3) and their hydrographs are similar and expected to be type 2 and occurred as seen in Figure 4. The hydrograph shapes of the Kizlar and Egridere basins may be categorized similarly. The slopes of the hydrographs are sharply up to the time of peak and then become mild (Figure 4). The hydrographs of the Akdogan and Guvenc basins are also similar in shape. The slopes of rising and recession curves of hydrographs are almost equal. The Akdogan and Guvenc basins have the first 2 ranks for the relief ratio. The expected hydrograph shape and type are presented in Figure 1(A) and Figure 2 (type 1), respectively, and the hydrograph itself is in Figure 4. In the case of low drainage density, the hydrograph is expected to be type 2 (Figure 2). The Kizlar and Guvenc basins have the lowest drainage densities among the 5 basins (Table 2), but they have the highest peak values (Figure 4). This is not a common case for general applications. The possible reason might be that some factors (vegetation, soil properties etc.) affect runoff. A qualitative comparison (Chi-square tests) indicated that the algorithm used to quantify the basin characteristics was valid and the differences in the quantification of characteristics were not significant (Table 2). Chi-square tests showed that the most significant difference between the manual and algorithm calculations was observed in the total stream length (1.055). The next significant difference was in the contour length (0.387). The Chi-square values for slope, drainage density, and median elevation were greater than 0.1, but these values were quite small for the remaining 11 parameters. However, since Chi-square values of all parameters were less than 9.488, the algorithm method can be used in this type of study. Since all basin characteristics were determined using four vector coverages, digitization should be carefully done. If basin characteristics are very important for any study and the basin area is relatively small, an available larger scale map (i.e. 1/5000) should be used to increase the accuracy. If the basin area is large and a map with a scale of 1/5000 is not available, then 1/25 000 scale may be used. By increasing the scale, the accuracy is increased; consequently curves of mainstream and branches are more precise and basin characteristics are determined more accurately. 139 140 H. Apaydin et al. Downloaded from http://iwaponline.com/hr/article-pdf/37/2/129/364736/129.pdf by guest Table 2 The results of manual and algorithm methods Basins Basin characteristics x 2 * Akdogan Area (km2) Perimeter (km) Main stream length (km) Total stream length (km) Contour length (km) Form factor Circularity ratio Compactness ratio Basin elongation Slope (%) (Williams and Berndt 1972) Slope (%) (Wisler and Brater 1959) Drainage density (km/km2) Maximum relief (m) Relief ratio Relative relief (%) Eff. basin width (km) Median elevation (m) * x2: Chi-square results Egridere Guvenc Kizlar Kocacesme Manual Algorithm Manual Algorithm Manual Algorithm Manual Algorithm Manual Algorithm 7.38 11.15 4.80 14.50 22.97 0.34 0.75 1.16 0.66 15.06 15.26 1.96 406 0.088 3.64 1.59 1132.0 7.57 11.74 4.59 15.12 23.09 0.36 0.69 1.20 0.68 15.04 15.24 2.00 406 0.088 3.46 1.65 1123.4 10.49 14.25 6.85 34.35 53.75 0.28 0.65 1.24 0.59 24.36 27.52 3.27 350 0.057 2.46 1.71 320.20 10.87 15.01 7.01 38.14 55.52 0.22 0.61 1.28 0.53 23.68 25.53 3.51 350 0.050 2.33 1.55 318.13 16.13 19.00 5.40 35.50 64.88 0.69 0.56 1.33 0.93 20.01 20.12 2.20 406 0.083 2.14 3.33 1238.45 16.12 18.76 5.13 31.51 65.49 0.61 0.58 1.32 0.88 19.43 20.31 1.95 406 0.079 2.17 3.14 1236.82 26.25 21.50 6.30 39.00 43.75 0.87 0.71 1.18 1.05 7.41 8.17 1.49 269 0.049 1.25 4.77 702.00 27.02 21.93 6.10 40.43 45.39 0.69 0.71 1.19 0.94 7.63 8.40 1.87 269 0.043 1.23 4.32 699.32 11.30 16.25 6.50 27.00 31.13 0.27 0.54 1.36 0.58 15.29 15.71 2.39 448 0.069 2.76 1.73 1208.34 11.35 16.57 6.88 28.67 34.14 0.24 0.52 1.39 0.55 13.92 15.04 2.53 448 0.065 2.70 1.65 1211.45 0.040 0.086 0.055 1.055 0.387 0.079 0.009 0.003 0.025 0.178 0.193 0.134 0.000 0.002 0.019 0.081 0.100 Although the developed algorithm uses vector data, it can be used for raster data. If digital elevation model of the basin is available, all four coverages can also be extracted using grid elevation data. Most of GIS and RS software have a capacity for determination of basin boundary and flow path (stream and branch). These raster data can be converted to vector data. Additionally, contour lines can be extracted from elevation data. In this case; resolution of raster data has an important impact on the results. The application of ecological principles to basin planning has recently become one of the most important topics of natural resources management discussions. Effective management of a basin depends on a comprehensive understanding of the components of basins and their interaction. To understand the interrelationships in morphological systems and in process response systems, it is necessary to express the character of the drainage basin in quantitative terms and correct rational equations (Gregory and Walling 1973; Reimold 1998). This paper presents an Arcview script for determining the basin characteristics. A computer program was written to determine basin area, stream length, basin shape, form factor, circularity ratio, compactness ratio, basin elongation, basin slope, drainage density, relief, basin width and median elevation. The developed method easily offers a direct determination of these quantitative terms. The most significant difference between the manual and algorithm methods was observed in the calculation of total stream length. The difference between these two methods was relatively significant in calculation of contour length, slope, drainage density, and median elevation, while this difference was quite small for the remaining 11 parameters. Compared to manual methods of measurement, the algorithm significantly decreases the amount of time and effort required to quantify selected characteristics for basins, particularly when a large number of drainage basins should be processed. Still, the geomorphologic approaches have some validity and should not be abandoned in future research (Ritter et al. 1995). H. Apaydin et al. Conclusions References Aronoff, S. (1989). Geographic Information Systems: A Management Perspective, WDL, Ottowa. Aykanli, N., Acar, C.O. and Yolcu, G. (2002). Canakkale-Bayramic-Egridere basin precipitation and runoff research report (in Turkish). Soil and Water Resources Researchs, KHGM Publications, Ankara, Turkey, 16 – 31. Berger, K.P. and Entekhabi, D. (2001). Basin hydrologic response relations to distributed physiographic descriptors and climate. J. Hydrol., 247, 169 – 182. Bhaskar, N.R., James, W.P. and Devulapalli, R.S. (1992). Hydrologic parameter estimation using geographic information system. J. Wat. Res. Plann. Mngmnt., ASCE, 118(5), 492 – 512. Black, P.E. (1991). Watershed Hydrology, Prentice Hall, Englewood Cliffs, New Jersey. Carlston, C.W. Drainage density and streamflow. USGS Prof. Paper 422-C. Costa, J.E. (1987). Hydraulics and basin morphometry of the largest flash floods in the conterminous United States. J. Hydrol., 93, 313 – 338. Chow, V.T. (1964). Handbook of Applied Hydrology, McGraw-Hill, New York. Doan, J.H. (2000). Development and application of geospatial hydrologic model extension HEC-GeoHMS. ESRI User Conference, ESRI, Redlands, p. 692. Dehn, M., Gartner, H. and Dikau, R. (2001). Principles of semantic of landform structures. Comput. Geosci., 27, 1005– 1010. DeVantier, B.A. and Feldman, A.D. (1993). Review of GIS applications in hydrologic modeling. J. Wat. Res. Plann. Mngmnt., ASCE, 119(2), 246 – 261. 141 Downloaded from http://iwaponline.com/hr/article-pdf/37/2/129/364736/129.pdf by guest H. Apaydin et al. Eash, D.A. (1993a). A geographic information system procedure to quantify physical basin characteristics. In: Proc. of the Symposium on Geographic Information Systems and Water Resources, AWRA, Bethesda, MD, 173 – 182. Eash, D.A. (1993b). Estimating design-flood discharges for streams in Iowa using drainage-basin and channelgeometry characteristics. Water-Resources Investigations Report 93-4062, US Geological Survey, IA. ESRI (1994). Introduction to ArcView, ESRI, Redlands, CA. ESRI (1995). Programming with Avenue, ESRI, Redlands, CA. ESRI (2001). ArcInfo Users Guide, ver 8.1.1. ESRI, Redlands, CA. FAO (1988). Geographic Information Systems in FAO, Food and Agriculture Organization of the United Nations, Rome. Glymph, L.M. and Holtan, H.N. (1969). Land treatment in agricultural watershed hydrology research. Effects of Watershed Change on Streamflow, Univ. of Texas Press, Texas, 44 –68. Gregory, K.J. and Walling, D.E. (1973). Drainage Basin Form and Process, Halsted, New York. Harvey, C.A. and Eash, D.A. (1996). Description, instructions, and verification for Basinsoft, a computer program to quantify drainage-basin characteristics. Water-Resources Investigations Report 93-4287, US Geological Survey, IA. Horton, R.E. (1932). Drainage basin characteristics. Trans. Am. Geophys. Union, 13, 350 – 361. Horton, R.E. (1941). Sheet erosion: present and past. Trans. Am. Geophys. Union, 22, 299 –305. Jenson, S.K. and Domingue, J.O. (1988). Extracting topographic structure from digital elevation data for geographical information system analysis. Photogramm. Engng. Remote Sensing, 54(11), 1593– 1600. Karas, E. (2001). Kutahya-Kocacesme basin precipitation and runoff research report (in Turkish). Soil and Water Resources Researchs, KHGM Publications, Ankara, Turkey, 124 – 138. Kaya, S. and Helaloglu, C. (2003). Sanliurfa-Kizlar basin precipitation and runoff research report (in Turkish). Soil and Water Resources Researchs, KHGM Publications, Ankara, Turkey, 110 –127. Lacroix, M.P., Martz, L.W., Kite, G.W. and Garbrecht, J. (2002). Using digital terrain analysis modeling techniques for the parameterization of a hydrologic model. Environ. Modell. Software, 17, 127 – 136. Linsley, R.K., Kohler, M.A. and Paulhus, L.H. (1988). Hydrology for Engineers, McGraw-Hill, London. Mohamoud, Y. (2004). Comparison of Hydrologic Responses at Different Watershed Scales, US Environmental Protection Agency, NC. Morisawa, M.E. (1967). Relation of discharge and stream length in eastern United States. In: Proc. Int. Hyd. Symp., Fort Collins, CO, Colorado State University, Colorado, 173 – 176. Oguz, I. and Balcin, M. (2003). Tokat-Zile-Akdogan basin precipitation and runoff research report (in Turkish). In: Soil and Water Resources Researchs, KHGM Publications, Ankara, Turkey, 64 – 78. Patton, P.C. and Baker, V.R. (1976). Morphometry and floods in small drainage basins subject to diverse hydrogeomorphic controls. Wat. Res. Res., 12, 941 – 952. Reimold, R.J. (1998). Watershed Management. Practice, Policies, and Coordination, McGraw-Hill, New York. Ries, K.G. (2002). STREAMSTATS: a US Geological Survey web site for stream information. In: Hydroinformatics 2002 Conference, Cardiff, Wales, July, IWA Publishing, London, pp. 1– 5. Ritter, D.F., Kochel, R.C. and Miller, J.R. (1995). Process Geomorphology, Wm. C. Brown, New York. Sankarasubramanian, A. and Vogel, R.M. (2002). Comment on the paper: “Basin hydrologic response relations to distributed physiographic descriptors and climate” by Berger, K.P. and Entekhabi, D. 2001. J. Hydrol., 247:169 – 182. J. Hydrol. 263, 257 –261. Schumm, S. (1954). The relation of drainage basin relief to sediment loss. In: Int. Union Geodesy Geophys. 10th Gen. Assembly. Int. Assoc. Sci. Hydrol. Publ. 36(1), 216 – 219. Sefton, C.E.M. and Howarth, S.M. (1998). Relationships between response characteristics and physical descriptors of catchments in England and Wales. J. Hydrol., 211, 1 – 16. Spence, C., Dalton, A. and Kite, G.W. (1995). GIS supports hydrological modeling. GIS World, 8(1), 62 –65. Tekeli, Y.I. and Babayigit, H.G. (2003). Ankara-Yenimahalle-Guvenc basin precipitation and runoff research report (in Turkish). Soil and Water Resources Researchs, KHGM Publications, Ankara, Turkey, pp 33 –48. Ward, A.D. and Trimble, S.W. (2004). Environmental Hydrology, 2nd edn. Lewis Publishers, Florida. Williams, J.R. and Berndt, H.D. (1972). Sediment yield computed with Universal Equation. Proc. Am. Soc. Civil Engrs., 98, HY12. Wisler, C.O. and Brater, E.F. (1959). Hydrology, John Wiley and Sons, New York. 142 Downloaded from http://iwaponline.com/hr/article-pdf/37/2/129/364736/129.pdf by guest