Currently, sampling regimes rely on predetermined time or flow-weighted intervals at which to collect samples. During periods of high flow, such as storm events, automated samplers may run out of sample bottles before the end of the hydrograph, resulting in the loss of data and inaccurate load calculations. In this research, real-time monitoring sensor networks with automated samplers were used to monitor the flowrate in agricultural tile-drains and ditches. By characterizing site specific hydrographs, a model was developed that enabled the datalogger to create storm-specific sampling regimes, resulting in approximately 20 samples for each storm event that resulted in a significant hydrograph. The model uses a power-law relationship for defining each hydrograph’s recession curve and spaces sampling events at flow-weighted intervals over the expected recession. The model’s performance was found to be satisfactory when either average or best-fit values of the two model parameters were used to predict the recession curves. The average error in the cumulative flow represented by each recession sample in the six hydrographs analyzed was less than 10%. The use of this model to create a storm-specific sampling regime eliminated the uncertainty in determining the flow-weighted sampling interval and led to the optimal usage of the sample bottles available in the automated sampler.