Limitations to measuring riverine denitrification at the whole reach scale: effects of channel geometry, wind velocity, sampling interval, and temperature inputs of N2-enriched groundwater

Abstract

A model-based approach was recently introduced for measuring riverine denitrification based on measured changes in dissolved N2 concentration during riverine transport (Laursen & Seitzinger, 2002a). Inputs to the model, including water temperature, channel depth, wind velocity, and time-of-travel between sampling locations, vary greatly among natural systems. Simulations were run by varying the values of these inputs and determining rates of N2 accumulation in river water and the detection limits for measuring denitrification using this method. Dinitrogen was found to accumulate most rapidly in streams that were shallow, particularly under conditions of low wind velocity. Dissolved N2 concentrations, modeled in rivers with a diurnal temperature variation of 5 � C and under conditions of no denitrification or 1 mmol Nm )2 h )1 , showed that sensitivity of the method can vary as temperatures change. Under low wind conditions and in rivers <1m in depth, this method is capable of detecting denitrification rates as low as 30–100 lmol N m )2 h )1 . This limit of detection should be adequate to measure in situ rates in many North American streams, particularly in agricultural watersheds. In deeper rivers N2 accumulated more slowly and the method became less sensitive. The results of this study should guide decisions regarding the application of this method based on the specific characteristics of a study reach (channel geometry) and the physical conditions (i.e. wind velocity and water temperature) under which measurements are to be made. The input of N2-enriched groundwater along a study reach can result in N2 accumulation that could be misinterpreted as denitrification. Some knowledge of the inputs of groundwater along a reach should also guide decisions regarding the application of this method.