Comparing the subsurface drainage flow prediction of the DRAINMOD and ADAPT models for a cold climate

The DRAINMOD computer model has been widely used for simulating the performance of subsurface drainage systems. The ADAPT model was created by merging components of DRAINMOD and GLEAMS and has evolved for over ten years, including the recent addition of soil freeze/thaw processes. DRAINMOD was also recently modified for soil freeze/thaw processes for application in cold climates. Computational time step, method of ET estimation, and soil freeze/ thaw processes are examples of ways in which current versions of these models differ from one another. Previous comparisons of these models were made for warmer climates, before the addition of cold-climate hydrology to both models. The performances of DRAINMOD and ADAPT were compared for cold-climate conditions, calibrated using two years of observed data from a 23-ha farm field in southern Minnesota. Model performance was evaluated and compared for seasonal, monthly, daily, and event-based time scales and during snowmelt runoff periods. Observed data showed that 60% and 20% of annual subsurface drainage runoff occurred during the transition period between winter and spring (snowmelt period) in 1998 and 1999, respectively. DRAINMOD overpredicted drainage by 11% and 25% for these periods, and ADAPT's results were within 10% of observed values for the snowmelt periods. Both models performed well at simulating the number and timing of drainage events in both snowmelt and later-season periods. The models performed best on a cumulative basis over the 2-year simulation period, where DRAINMOD overpredicted cumulative subsurface drainage by 1.7%, and ADAPT underestimated cumulative drainage by 0.2%. The models diverged in their abilities to predict the largest daily drainage events: DRAINMOD overpredicted and ADAPT underpredicted these events. Substantially more effort was required to calibrate ADAPT because of the increased complexity of the model.