作者: S. Ferrant , C. Laplanche , G. Durbe , A. Probst , P. Dugast
DOI: 10.1002/HYP.9324
关键词:
摘要: A nitrate sensor has been set up to measure every 10 min the signal in a stream draining small agricultural catchment dominated by fertilized crops during 2-year study period (2006–2008) south-west of France. An situ sampling protocol using automatic sampler monitor flood events have used assume point-to-point calibration values. The concentration exhibits nonsystematic and dilution effects events. We demonstrate that calibrated gathered from outlet is considered be continuous Nyquist–Shannon theorem. objectives this are quantify errors generated typical infrequent design appropriate strategy according objectives. Nitrate flow data numerically sampled simulate common frequencies. total fluxes calculated simulated samples compared with reference value computed on signal. Uncertainties increasing as intervals increase; method not discharge compute bring larger uncertainty. dispersion bias for each interval evaluate uncertainty hydrological period. High underestimation made periods when high-concentration overlooked. On contrary, high frequencies (from 3 h 1 day) lead systematic overestimation (bias around 3%): highest concentrations overweighted interpolation such case. generates less than 1% load estimation error sample peaks. consider useful newly emerging field technologies assess short-term variations water quality parameters, minimize number analysed state at any time.