UV-VIS Absorbance time series forecasting for water quality in a constructed wetland using ARIMA

When discussing urban drainage systems, water quality prediction plays a crucial role, given that integrated management of this resource is required to meet human needs. The present paper applies ARIMA (Autoregressive Integrated Moving Average) to forecast affluent and effluent water quality, in addition to the contaminant removal efficiency, of a constructed wetland located on the campus of the Pontificia Universidad Javeriana in Bogotá, Colombia. Measurements for the ARIMA predictions rely on time series obtained by UV-Vis spectrometry probes. ARIMA-based predictions appropriately forecast the first 12 hours of the water quality time series for the three data sets analyzed: affluent, effluent and efficiency. Overall prediction errors did not exceed 15% for any of the observed data. Separate analyses of affluent and effluent aver that relative forecast errors resulting from ARIMA prove to be less significant for UV wavelengths than for the visible (Vis) range. Likewise, for the UV range, this type of error demonstrates less variability than it does for the Vis range, a result which suggests that ARIMA is a worthwhile prediction method when discussing contaminants that fall in the UV range.