Machine learning approach towards explaining water quality dynamics in an urbanised river

Benjamin Schäfer, Christian Beck, Hefin Rhys, Helena Soteriou, Paul Jennings, Allen Beechey, Catherine M. Heppell
Human activities alter river water quality and quantity, with consequences for the ecosystems of urbanised rivers. Quantifying the role of human-induced drivers in controlling spatio-temporal patterns in water quality is critical to develop successful strategies for improving the ecological health of urban rivers. Here, we analyse high-frequency electrical conductivity and temperature data collected from the River Chess in South-East England during a Citizen Science project. Utilizing machine
more » ... arning, we find that boosted trees outperform GAM and accurately describe water quality dynamics with less than 1% error. SHapley Additive exPlanations reveal the importance of and the (inter)dependencies between the individual variables, such as river level and Wastewater Treatment Works (WWTW) outflow. WWTW outflows give rise to diurnal variations in electrical conductivity, which are detectable throughout the year, and to an increase in average water temperature of 1 o C in a 2 km reach downstream of the wastewater treatment works during low flows. Overall, we showcase how high-frequency water quality measurements initiated by a Citizen Science project, together with machine learning techniques, can help untangle key drivers of water quality dynamics in an urbanised chalk stream. Across the globe human activities, such as urbanisation, are causing changes to catchment water cycles that have profound impacts on the water quantity, quality and the ecology of rivers 1,2 . Urbanisation changes the hydrology of a catchment in multiple ways 3 . The introduction of impermeable surfaces, together with artificial drainage systems, can increase peak flows in rivers 3 , reduce hydrological response times to rainfall 4,5 and reduce baseflow and groundwater recharge 6 . In many countries combined sewer systems carry domestic and industrial wastewater to wastewater treatment plants under dry conditions, along with stormwater drainage from paved areas when it rains. Treated effluent discharge can cause distinctive flow patterns in rivers dictated by human activity 7, 8 . The human activities associated with urbanisation also have impacts on chemical water quality, with lots of effort focused on characterising water quality changes in urban rivers during storm events [9] [10] [11] , and the 'first flush' phenomenon 12, 13 . Combined sewer overflows can also significantly impact water quality following intense rainfall events when the capacity of wastewater treatment works has been exceeded [14] [15] [16] [17] . We also know that urban streams tend to have higher mean electrical conductivity and major ion concentrations in comparison to their rural counterparts [18] [19] [20] , which arises from a combination of point and diffuse pollution sources. For example, chloride, sulphate, sodium and potassium are common electrolytes in urine and therefore concentrated in wastewater 19 . Determining the main sources of individual ions in urban systems, however, has proven to be challenging 21 . Such elevated solute levels are now leading ecologists to hypothesise about potential implications of elevated ionic concentrations for the health and resilience of urban stream ecosystems 22 . Human activities are causing widespread degradation of water quality in rivers with consequences for ecological health 23 . These activities lead to changes to the water quality of receiving waters which operate over nested timescales from hours (in response to rainfall events) to daily, seasonal and inter-annual cycles 24, 25 . To fully understand the impact of human activities on rivers we need to tease apart human-induced and natural variations
doi:10.5445/ir/1000149444 fatcat:66yjulfdqfczrh552iupmozame