Classification of Prairie basins by their hysteretic connected functions

Diagnosing climate change impacts in the post-glacial landscapes of the North American Prairies through hydrological modelling is made difficult by drainage basin physiography. The region is cold, dry and flat with poorly developed stream networks, and so the basin area that is hydrologically connected to the stream outlet varies with basin depressional storage. The connected area controls the contributing area for runoff reaching the stream outlet. As depressional storage fills, ponds spill from one to another; the chain of spilling ponds allows water to flow over the landscape and increases the connected area of the basin. As depressional storage decreases, the connected fraction drops dramatically. Detailed, fine-scale models and remote sensing have shown that the relationship between connected area and the depressional storage is hysteretic in Prairie basins and that the nature of hysteresis varies with basin physiography. This hysteresis needs to be represented in hydrological models to calculate contributing area, and therefore streamflow hydrographs. Parameterisations of the hysteresis are needed for large-scale models used for climate change diagnosis. However, use of parameterisations of hysteresis requires guidance on how to represent them for a particular basin. This study shows that it is possible to relate the shape of hysteretic functions as determined by detailed models to the overall physiography of the basin, such as the fraction of the basin below the outlet, and remote sensing estimates of depressional storage, using the size distribution and location of maximum ponded water areas. By classifying basin physiography, the hysteresis of connected area - storage relationships can be estimated for basins that do not have high-resolution topographic data, and without computationally-expensive high-resolution modelling.