U.S. Geological Survey
U.S. GEOLOGICAL SURVEY BULLETIN 2016
Selected Papers in the Applied Computer Sciences 1992

CHAPTER D

Integrating and Aggregating Data From Stream-Drainage Systems Using a Geographic Information System

By David L. Lorenz

Prepared in cooperation with the Minnesota Department of Transportation

CONTENTS

Abstract
Introduction
Stream-Drainage Integration
Assumptions
Methods
Special Conditions
Stream-Drainage Aggregation
Summary
Reference Cited

FIGURES

1-3. Maps showing:
1. Hypothetical drainage basin (65kb)

2. A simple stream-drainage system (64kb)

3. A complex stream-drainage system (59kb)

TABLES

1. Selected data for subbasins for simple stream-drainage system shown in figure 2 (4kb)

2. Selected data from stream segments for simple stream-drainage system shown in figure 2 (6kb)

3. Selected data for subbasins for complex stream-drainage system shown in figure 3 (4kb)

4. Selected data from stream segments for complex stream-drainage system shown in figure 3 (7k)

5. Selected aggregated data for complex stream-drainage system shown in figure 3 (9kb)

Abstract

Data from streams and drainage basins are recorded, integrated, and analyzed using a geographic information system. The watercourse and drainage-basin data bases include geographic, attribute, and supplemental data. The geographic information system has f unctions to associate the watercourse data with the drainage-basin data. A computer program is used to process and integrate data to describe the stream-drainage system. The data base and program provide a framework to automate the calculation of aggregat ed basin characteristics (such as drainage area or lake area) and stream characteristics (such as main-channel length) for any delineated basin.

INTRODUCTION

The U.S. Geological Survey (USGS), in cooperation with the Minnesota Department of Transportation, is determining selected characteristics of streams in the Minnesota River Basin. As part of this project, drainage basins are delineated upstream of USGS st reamflow-gaging stations, outfalls of municipal sewage-treatment plants, and mouths of major tributary streams. In general, the basins are delineated (at a scale of 1:24,000) for tributaries draining at least 5 square miles.

The watercourse system and drainage-basin boundaries are recorded in the ARC/INFO geographic information system (GIS). The GIS watercourse and drainage-basin data bases include geographic data (line segments defining both the watercourse and the basin bou ndaries), attribute data (attributes of the geographic data such as the basin in which a watercourse segment is located or the area of a basin), and supplemental data (information about watercourse segments or noncontributing basins).

ARC/INFO includes the capability to associate line segments with polygonal areas, to define the direction of flow, and to determine the distance between selected locations on a flow path. However, these capabilities of the GIS are insufficient to fully in tegrate stream systems with drainage basins. Therefore, two programs for determining all subbasins and stream segments upstream of the outlet of a drainage basin were developed. The programs calculate the drainage area of the basin, the lake area and tota l storage within the basin, and the main-channel length of the watercourse upstream of the outlet of the basin.

A FORTRAN computer program is used to process and integrate the data to describe the stream-drainage system. An INFO program automates the calculation of aggregated basin characteristics (such as drainage area or lake area). These programs run within the ARC/INFO GIS on a PRIME series 9955 computer.

This report describes the computer programs and the assumptions required for the execution of the programs.

STREAM-DRAINAGE INTEGRATION

The stream-drainage integration program is written in FORTRAN and executed from an ARC macro language (AML) program. The integration program was written to be easily portable to other computer systems running ARC/INFO but needs slight modification to run on another computer system.

This section of the report describes the assumptions about the data needed to run the integration program and a brief discussion of the method used by the program to integrate stream-drainage data. Included also are discussions of two special cases that t he program can handle.

Assumptions

The following assumptions control the definitions of the watercourse data and drainage-basin data within the context of the GIS:
  1. Line segments are the graphical entities necessary to represent basin boundaries and watercourses. Linear features can be made from several segments, but segments must have endpoints where segments contact each other. Line segments can be joined together to represent a linear feature (watercourse) or form a closed area called a polygon (subbasin).
  2. Subbasins are represented by polygons that are uniquely identified by the GIS.
  3. Subbasin boundaries can be delineated beginning at any point on a stream. The boundaries typically begin at the confluence of major tributaries, streamflow gaging stations, sewage-treatment-plant outfalls, and mouths of major tributaries to lakes.
  4. Subbasins that do not contribute directly to overland runoff be delineated and identified by the user.
  5. The watercourse data structure is a dichotomous, non-braided, nondivergent, connected, dendritic representation of the stream system.
  6. An individual watercourse is defined as a single line from its confluence or mouth to the basin divide. At each branch of the stream, the line follows the branch draining the largest area. The line is continuous and passes through marshes, lakes, and the midline of wide rivers and braided streams. The watercourse has a unique, arbitrary identifier assigned by the integration program.
  7. The segments of the watercourse that lie within a subbasin also have an identifier assigned by the GIS that associates the segment with the subbasin.

Methods

The integration program makes four passes through the stream data to complete the process and determine contributing basins. The program checks that the data satisfy the assumptions and calculates all of the necessary output data. The first pass proceeds upstream from the most downstream segment, checking for dichotomous branching, the from-to direction of the segment, and the segment number of the downstream segment. The second pass proceeds downstream from the most upstream segments and determines the c umulative area associated with each segment and its Strahler number (Gregory and Walling, 1973, p. 43). The third pass proceeds upstream from the most downstream segment, following the segments that drain the largest area and recording their arbitrary str eam identifiers. The final pass proceeds downstream from the most upstream segment of each stream and determines the modified-Horton-order number (Gregory and Walling, 1973, p. 43) for the stream and a different arbitrary downstream-order number for each subbasin. The range of downstream-order numbers that contribute flow to the basin also is determined during the final pass.

A hypothetical drainage basin containing streams, lakes, and marsh areas is shown in figure 1. An example of how the basin in figure 1 may be divided by the user into subbasins is shown in figure 2. The example shows the segments representing the watercou rses and their respective segment numbers and the subbasin boundaries and their respective basin numbers. Lists of the data generated by the integration program for the example in figure 2 are contained in tables 1 and 2.

The downstream-order number in table 1 is an arbitrary number assigned to a subbasin. The number increases from the uppermost subbasin on the main stream down to the basin outlet. Downstream-order numbers are dependent on the number of watercourse segment s and noncontributing basins and, therefore, not necessarily sequential. The number also increases from the uppermost subbasin of a tributary down to the tributary outlet. This way of assigning numbers to basins provides a technique for easily aggregating data for basins because only those subbasins that contribute flow have a downstream-order number between numbers for the uppermost subbasin and the outlet subbasin.

The stream identifier in table 2 is an arbitrary number assigned to the watercourse that drains the larger area. This number defines which segments are selected to calculate stream characteristics such as main-channel length.

The cumulative area of a segment is defined as the total area of the drainage subbasins upstream of the downstream end of the segment. If the subbasin has many segments, the area of the subbasin is distributed proportionally according to the lengths of th e segments within that subbasin.

The modified-Horton-order number is a description of the morphology of a stream system. A modified-Horton-order number of 1 designates the main stream of the system. Tributaries to the main stream have modified-Horton-order numbers of 2, and so forth.

The Strahler number is another description of the morphology of the stream system. A Strahler number of 1 indicates that the stream is the uppermost tributary. The Strahler number increases by 1 at the confluence of two streams with equal Strahler numbers . The Strahler number is the greater of the two where two streams of unequal Strahler numbers join.

Special Conditions

Delineating subbasins is not always possible in the simple approach shown in figure 2. Some subbasins do not directly contribute runoff to the stream system; and furthermore, subbasins cannot always be delineated from the confluence of two streams. Figure 3 shows an example of a subbasin that does not contribute runoff (subbasin number 8) and a subbasin that is delineated above the confluence of the watercourse (subbasin number 9).

A subbasin that does not normally contribute runoff to the stream system, called a noncontributing subbasin, can be incorporated into the stream-drainage system by assigning the subbasin to a segment of a watercourse-or a noncontributing subbasin can be o mitted. The integration program checks each segment or noncontributing subbasins that have been assigned to it. If a segment has a noncontributing subbasin assigned to it, the area of that subbasin is added to the cumulative area for that segment. Noncon tributing subbasins can be omitted by not assigning the area to a segment; the cumulative area then represents only subbasins contributing flow to the stream system. Subbasin number 8 was assigned to stream-segment number 6 in figure 3.

Subbasins delineated upstream of the junction of the watercourses, called isolated subbasins, occur commonly for large tributaries to lakes; for example, subbasin 9 in figure 3. Segment number 11 does not properly belong to subbasin 6, and the cumulative area for segment 11 must represent only the drainage area upstream of the outlet of subbasin 9. The integration program checks each segment to determine if the area of the subbasin should not be accumulated. The stream system must be checked by the user for segments that drain isolated subbasins.

Lists of data for the example of the noncontributing subbasin and the isolated subbasin shown in figure 3 are contained in tables 3 and 4.

The method of distributing the subbasin area according to segment length can produce errors where a large part of the subbasin drains into a short segment. Segment number 10 in figure 3 is much shorter than segment number 8 and should be allocated a larger portion of the subbasin area. Basins with many tributaries must be checked by the user to determine if the distribution of subbasin area has affected the stream-numbering system.

STREAM-DRAINAGE AGGREGATION

After the watercourse and drainage-basin data are integrated, a second program is run to aggregate the data and calculate the drainage area, lake area, and total storage area. This program was written in the INFO programming language and must be changed w hen the basin and stream data base names change. No other changes should be necessary when changing computer systems.

The Strahler and modified-Horton-order numbers are by-products of the integration program and are not used by the aggregation program. These numbers do characterize the morphology of the stream, however, and could be used in a hydrological analysis that incorporates stream morphology. The aggregation program requires the downstream-order number and the uppermost downstream-order numbers, selects those subbasins in that range (for each subbasin), and calculates the drainage characteristics for each subbasi n.

The aggregation program requires that the polygons representing lakes and marsh areas be associated with a subbasin. The GIS performs this association and totals the areas within each subbasin. The program steps through the data one subbasin at a time, ca lculates the percentage of area covered by lake and marsh areas within the basin, selects all the subbasins contributing flow to that subbasin, and calculates the cumulative drainage area and the percentage of area covered by lake and marsh areas. A list of drainage area, percentage lake area, and percentage storage area for the subbasins in figure 3 is contained in table 5.

SUMMARY

Data from streams and drainage basins are recorded, integrated, and analyzed using a geographic information system. The watercourse and drainage-basin data bases includes geographic, attribute, and supplemental data. The geographic information system has functions to associate the watercourse data with the drainage-basin data. A program is run to determine the subbasins contributing flow to the outlet of each basin. Another program produces a table summarizing the area, lake, total-storage, and stream dat a for each subbasin.

REFERENCE CITED

Gregory, K.J., and Walling,D.E., 1973,
Drainage basin form and process: New York, John Wiley, 456 p


Return to Chapter C

Continue to Chapter E

Return to Bulletin 2016 contents

Return to USGS Home Page


U.S. Geological Survey, ISD National Center, Reston, VA 22092, USA
URL http://pubs.usgs.gov/bulletin/b2016/chapa/ch_a.html
Contact: webmaster@pubs.usgs.gov
Last Modified: 9/8/95 (hem)