Crawford 1954; Buchanan and Somers 1969) and was devised in response to the .... that stream discharge had remained constant during the 3-hr period of field.
A Simple and Inexpensive Method for Determining Stream Discharge from a Streambank Daniel P. Molloy and Robert H. Struble Biological Survey New York State Museum State Education Department Cultural Education Center Albany, New York 12230
ABSTRACT A detailed protocol for the determination of stream discharge is presented which requires neither a current meter nor actual entry into stream water. 1be protocol represents a modification of the flotation method of discharge determination and was devised in response to the need for a simple, inexpensive, "bank-side" technique for estimation of stream flow. Discharge is determined tfirough measurements of a stream's crosssectional area and surface velocity. In tests at high and low flows, mean discharges determined by the proposed flotation method were not significantly different from those calculated using a standardized, currentmeter method. 1bis suggested that the proposed flotation method (a 10-20 min procedure) could be used with reasonable accuracy to determine discharge in small to moderate-sized streams.
INTRODUCTION Stream investigations often require characterization of a number of abiotic parameters, including discharge (i.e., volume of water passing a point per unit of time). In the absence of an established gauging station, discharge must be calculated, usually by an individual entering a stream and taking measurements of depth and velocity with a current meter at selected intervals across the stream (Buchanan and Somers 1969). The present paper presents a detailed protocol for the determination of stream discharge which requires neither a current meter nor actual entry into streamwater. The protocol represents a modification of the flotation method of discharge determination (Robins and Crawford 1954; Buchanan and Somers 1969) and was devised in response to the need for a simple, inexpensive, "bank-side" method for the estimation of stream discharge in the Adirondack Mountains of New York. The proposed protocol would likely be of value to field biologists in a variety of stream projects in which a current meter is unavailable and/or it would be impractical or hazardous for an individual to enter and wade across a stream for the purpose of measuring discharge.
METHODS AND MATERIALS As with any method of discharge calculation, selection of the site is a critically important factor. Ideally, the site should be a section of stream with a uniform bottom slope (bank to bank) and which is straight (i.e., parallel- sided banks), free-flowing, and sheltered from wind. Emergent or overhanging vegetation, pools, bends, turbulence, wind gusts, highly irregular bottom profiles, and debris will contribute to inaccurate measurements. The following equipment is required:
Cross-String: A nylon string (3 mm x 6 m) which resists tangling ic; recommended. The string should be clearly marked at 0.2-m intervals with alternating watetproof colors and weighted at one end. Lead weights (85-200 g), such as those used for duck decoys or bait casting, work well. To facilitate handling, the string should be wrapped around a dowel, attaching the unweighted end first. 477 Journal of Freshwater Ecology, Volume 4, Number 4- December, 1988 Copyright © 1988 by Oikos Publishers, Inc.
Figure 1. Taking depth readings from a stream bank: using a telescoping pole and depth- and cross-strings.
Depth-String: Another nylon string (3 mrn x 2m) with an attached weight and clearly marked at 0.05-m intervals serves as the depth string. Intervals of 0.25-m can be distinguished with marks of another color. The string is attached to the pole. Pole: An extension pole, either telescoping or sectional for easy carrying, is needed. Aluminum golf ball retrievers ($1 0-40) work well and can extend up to 6 m. Determining Cross-sectional Area
From the streambank, an adequate amount of the cross-string is gently tossed (weighted end) to the opposite bank. The landing place of the weight on the opposite bank is not critical, provided it is not in the water. The string should be pulled taut and perpendicular to the stream flow and attached to a fixed object. When pulling the string taut, it is advisable to line-up one of the string's interval marks at the edge of the near streambank. The pole is extended, with the depth-string attached, so that it will be long enough to reach across the stream. The weighted end of the string should be suspended ca. 0.5 m greater than the estimated maximum depth of the water to be measured. Then 5-l 0 depth readings are taken at equal intervals from each other (determined by the cross-string) across the stream's width, beginning at a distance from the bank of one half the interval between measurements. (For example, if depth readings are made at 0.5 m intervals, the fust reading should be taken at 0.25 m from the stream bank, and the second reading taken at 0.75 m, etc.). Five to ten measurements are chosen for economy of time; accuracy would increase with more measurements, especially if the stream bottom profile is highly irregular (e.g., boulders, large crevices, vegetation, etc.). Depth measurements are made by extending the pole and string over the water and lowering the weight into the water to the point at which contact is made with the stream bottom. Generally, this is felt through the pole and further confirmed by slack in the string (Fig. 1). With the depth information recorded, the total cross-sectional area (A) can be calculated as: A=DxW
478
where:
Dis the mean of all depth readings W is the total stream width (bank to bank)
An alternative method for calculating total cross-sectional area, the "midpoint method of area determination" (Buchanan and Somers 1969), is to sum the individual rectangular areas surrounding the depth readings, i.e., multiply the individual depth readings by the distance between depth readings and sum these individual areas.
Determining Velocity The mean velocity of the water passing through the cross-sectional area is determined from an adaptation of the flotation method (Robins and Crawford 1954; Buchanan and Somers 1969). In this method, mean surface velocity is calculated and then converted into the mean velocity of flow throughout the entire vertical water column. Surface velocity is measured by timing (minimum of 5.0 sec) the travel of a surface float over a known distance. (Wooden coffee-stir-sticks are ideal as surface floats since they are biodegradable and thus normally would not have to be retrieved.) The stream is visually divided into thirds (i.e., left, middle, right) and an equal number of flotation trials is performed in each section. Three trials/section are recommended, with the mean surface velocity being the average of all nine measurements. Using the calibrated cross-string, a section of stream length is measured equidistant upstream and downstream from the point at , which the cross-sectional area was determined. A float is tossed into the water near the upstream limit (of this distance); when the float passes the "starting point," it will be moving as fast as the water. The time for the float to pass from the starting point to the downstream measured point is recorded. The accurate timing of these floats is of critical importance to discharge determination (e.g., a 1.0-sec error in a 5.0-sec passage eventually results in a ca. 20% miscalculation of discharge). The stream discharge (Q) in L/min is ,£alculated as follows: Q = K X X A X 60,000 where: K is a constant Vis the mean surface velocity (rn/sec) A is the cross-sectional area (m2) 60,000 is the factor for converting m3/sec to L/min
v
The constant (K) is introduced to convert the mean surface velocity to the mean velocity throughout the vertical water column. Since surface velocity is always greater than the mean velocity of flow in the vertical, the value of K is always < 1.0. K values ranging from 0.8 to 0.9 (depending primarily on streambed morphology) have been recommended for general usage in streams (Anderson 1937, Robins and Crawford 1954, Buchanan and Somers 1969).
Determining the Accuracy of the Flotation Method
.--
The accuracy of the proposed flotation method was assessed by comparing discharge determinations made using the proposed method with those calculated using a widelyaccepted, standardized, current-meter method, i.e., six-tenths-depth, with a Price pygmy meter attached to a top-setting rod (Buchanan and Somers 1969). Field trials were conducted in Camden Creek (Shushan, NY) - a third-order, cobble-bottom stream. For each of the two methods, discharge determinations were made at five separate locations (i.e., ten locations in total) within a 100-m section of stream which lacked any tributaries. To evaluate the accuracy of the proposed method at different flow rates, trials were conducted when Camden Creek had relatively high and low discharge levels (March and August 1985, respectively). In both situations, observation of water depth at initiation and completion of the trials indicated that stream discharge had remained constant during the 3-hr period of field
479
measurements. Ten and thirty depth readings, respectively, were recorded across the stream during each execution of the proposed flotation and current-meter method. The mean discharges determined by the two methods were confirmed for homogeneity of variance (Fmax test) and then evaluated for statistically significant difference (t-test).
RESULTS AND DISCUSSION Mean discharges determined by the current-meter and the flotation methods were not significantly different in either the low or high discharge trials (P
