Quantifying and mapping inundation regimes within a large river-floodplain ecosystem for ecological and management applications

Spatial information on the distribution of ecosystem patterns and processes can be a critical component of designing and implementing effective management programs in river-floodplain ecosystems. For example, translating how flood pulses detected within a stream gauge record are spatially manifested across a river-valley bottom can be used to evaluate whether the current distribution of physical conditions has the potential to support priority habitats or if intervention is needed to meet desired goals. The size and complexity of large river-floodplain systems can make mapping inundation dynamics a challenging task. We used a geospatial model to simulate 40 years (1972-2011) of daily surface-water inundation depths for 11,331 km(2) of the Upper Mississippi River System floodplain. We identified discrete inundation events at each 4-m x 4-m pixel in the model as sequential days of submergence. We then quantified and mapped four aspects of inundation regime - event frequency, duration, magnitude, and timing - for each pixel. The spatial distribution of inundation regime attributes varied within and among multiple levels of river organization, including navigation pools and geomorphic reaches, but only event timing exhibited a strong down-river trend. Non-linear relations among inundation attributes and their geospatial distributions likely reflect complex interactions among topographic, hydrologic, and anthropogenic constraints on flooding dynamics. Together, our results reveal spatial gradients in inundation dynamics not captured by hydrologic data alone. Characterizing such diversity in inundation dynamics is important for testing hypotheses about ecological processes, developing models of ecosystem functions, and informing management actions.