Identifying Controls on Patterns of Intermittent Streamflow in Three Streams of the American Southwest: A Geospatial Approach

Despite a rising interest in intermittent river systems, landscape influences on long-term wetting and drying patterns of streamflow are not well understood. There has been a significant increase in the presence of intermittent rivers worldwide due to climate change and subsequent increases in groundwater abstraction, and these effects are intensified in already arid regions such as the American Southwest. Consequently, the spatial extent of wet and dry reaches of Arizona’s Agua Fria River, Cienega Creek, and San Pedro River has been documented by citizen scientists for several years. Citizen science involves the use of trained members of the public for data collection, and the analysis of datasets produced from citizen science projects have become a huge asset to the scientific community. Here, we synthesize the most current data (1999–2016) to determine what stream and valley characteristics act as drivers for patterns of surface water flow. Geologic, geomorphic, and land cover characteristics of these rivers were analyzed via aerial imagery and Digital Elevation Models within ArcGIS 10.3 in conjunction with the Soil and Water Assessment Tool model. Principal Component Analysis was used in order to assess trends across sites. A set of landscape intermittency metrics was produced and then further analyzed using Multiple Linear Regression. We found that land cover had a significant (p-value < 0.01) positive correlation with reach average (i.e., the proportion of channel wet). Physical watershed and channel characteristics each had a negative correlation with both intermittency metrics (i.e., wet/dry status and reach average). However, their results were not significant to the 0.05 level. This study begins to shed light on the drivers of landscape intermittency patterns of desert streams and demonstrates the utility of citizen science in regard to the study of intermittent river systems