Author
|
Submitted to: Transactions of the ASAE
Publication Type: Peer Reviewed Journal Publication Acceptance Date: 10/15/2002 Publication Date: 3/15/2003 Citation: KING, K.W., HARMEL, R.D. CONSIDERATIONS IN SELECTING A WATER QUALITY SAMPLING STRATEGY. TRANSACTIONS OF THE AMERICAN SOCIETY OF AGRICULTURAL ENGINEERS. 2003. v. 46(1). p. 63-73. Interpretive Summary: Public awareness and concern of pollutants such as nutrients, pesticides, and sediment entering our water systems has prompted many government entities and action groups to initiate monitoring programs aimed at quantifying the amounts of these pollutants entering and being transported through the system of waterways. When designing and implementing monitoring programs, one attempts to select a sampling strategy that will best capture what is actually occurring. While there exists an infinite number of sampling strategies, those most commonly used generally revolve around collecting samples base on a discrete time or a discrete flow past a point. Our efforts focused on quantifying the tradeoffs between several of the most commonly used sampling strategies. Approximately 300 runoff events distributed across 87 watersheds in the U.S. were used in the study. Findings of the study indicated using a time based sampling scheme generally gave more reliable results than a flow weighted scheme due to the number of samples that would be taken. It was also shown that collecting samples at time increments less than or equal to 15 minutes provided the best estimates of pollutant loadings. Prior to implementing a monitoring program, one should also give consideration to watershed characteristics such as drainage area, slope, etc. Technical Abstract: At the core of monitoring programs are schemes that determine when and how samples are taken for estimating stream loadings. Quantification of the differences between these schemes has not been documented. An analytic approach was used to evaluate 45 common sampling strategies that included time (5, 10, 15, 30, 60, 120, 180, 300, and 360-min) and flow-stratified (2.5, 5.0, 7.5, 10.0, 12.5 and 15.0 mm) schemes using discrete and composite sampling approaches. 305 storm hydrographs from 87 different watersheds in the U.S. were coupled with 2 concentration graphs (a 100% positive correlation of concentration to flow and a 100% negative correlation) to estimate average bias values for each sampling strategy. The mean bias values for time-based sampling always increased with an increase in sampling time interval. With respect to time- weighted sampling, a positive correlated concentration graph generally resulted in under-prediction (positive bias) from the true load, while a negative correlated concentration always resulted in over-prediction (negative bias). With respect to flow-stratified sampling, the direction of bias was reversed from the time-weighted case. Flow-stratified sampling at intervals greater than 7.5 mm does not adequately capture concentration points around the peak. At the lowest flow interval used in this study (2.5 mm), the bias associated with each correlation case was significantly different from zero (alpha=0.05). Time discrete sampling schemes less than or equal to 15-min provided the only bias values not significantly different from zero (alpha=0.05). The findings indicate that prior to water quality monitoring, consideration should be given to the sampling strategy and its impact on load estimates. |
