Deployment of a High-Density, High-Frequency Sensor Network to Quantify Stream Thermal Response to Natural and Anthropogenic Inputs in a Highly Urbanized Watershed

In 2021, the State of Maryland began to consider setting its first stream temperature TMDL (total maximum daily load). Scientific, watershed-scale evaluation of stream temperature to assess tradeoffs between point and nonpoint sources of thermal pollution is viewed as being valuable to for further policy development. In particular, policymakers are concerned that various types of stormwater management facilities may have a disproportionate impact on stream thermal regime. To gain insight into the impacts of stormwater facilities versus land cover on stream thermal regime, in 2021-2022 we deployed Onset Tidbit MX2203 sensors every 100 m along 16 km of a stream network in a highly developed watershed in suburban Baltimore, MD. Data are recorded every 5 minutes, with continuous data collection intended for 2 years across all flow regimes. Sensors are secured to the stream bed through a system of rebar, snap hooks, and zip ties. Data are retrieved from the sensors via Bluetooth technology, where sensors need to be removed from the stream for download. Evaluation of initially-downloaded data has revealed dramatic spatiotemporal behavior of storm runoff as well as influences of groundwater on thermal responses. Heat pulses generated from piped areas draining stormflow can be tracked as they move downstream and the response times for heating and subsequent cooling throughout the stream network can be calculated. At times when the groundwater temperature detectably differs from that of surface water, pulses of groundwater can be observed pushing into the stream at the onset of a storm before an upstream heat wave passes through a stream reach. Particularly interesting are thermal signals at night, when heated storm runoff is due to conduction of heat from hot impervious surfaces and when canopy effects are not present. Statistical analysis is ongoing, as well as development of video renderings of mapped thermal signals superimposed on the stream network.