The Autobot-WQ: A portable, low-cost autosampler to provide new insight into urban spatio-temporal water quality dynamics

Urbanization and the increase in urban land cover are growing concerns associated with numerous negative impacts on surface water quality. Currently, many emerging contaminants are difficult to measure with no field deployable sensors currently available. Hence, discrete grab samples are required for subsequent laboratory analysis. To capture the spatiotemporal variability in pollution pulses, autosamplers can be used, but commercial offerings are both expensive and have a large footprint. This can be problematic in urban environments where there is a high density of point source inputs and risk of vandalism or theft. Here, we present a small and robust low-cost autosampler that is ideally suited for deployment in urban environments. The design is based on "off the shelf" open-source hardware components and software and requires no prior engineering, electronics, or computer programming experience to build. The autosampler uses a small peristaltic pump to enable collection of 14 small volume samples (50 mL) and is housed in a small footprint camera case. To illustrate the technology, we present two use cases for rapid sampling of stormwater pulses of: 1) an urban river channel and 2) green roof runoff. When compared with a commercial autosampler, our device showed comparable results and enabled us to capture temporal dynamics in key water quality parameters (e.g., dissolved organic matter) following rain events in an urban stream. Water quality differences associated with differing green roof design/maintenance regimes (managed and unmanaged vegetation) were captured using the autosampler, highlighting how unmanaged vegetation has a greater potential for mitigating the rapid runoff and peaked pollutant inputs associated with impervious surfaces. These two case studies show that our portable autosampler provides capacity to improve understanding of the impact of urban design and infrastructure on water quality and can lead to the development of more effective mitigation solutions. Finally, we discuss opportunities for further technical refinement of our autosampler and applications to improve environmental monitoring. We propose a holistic monitoring approach to address some of the outstanding challenges in urban areas and enable monitoring to shift from discrete point sources towards characterization of catchment or network scale dynamics.