Statistical Modeling of Algae Concentrations in the Elbe River in the Years 1985—2001 Using Observations of Daily Dissolved Oxygen, Temperature, and pH

Abstract

AbstractThe aim of the paper is the reconstruction of daily algae concentrations (chlorophyll‐a) in the middle stretches of the Elbe River (Schnackenburg station) during the years 1985—2001. These data are required to verify changes in biogeochemical processes before and after the reunification of Germany. Both linear regression and multiple input single output (MISO) stochastic transfer function (STF) approaches are used in data reconstruction. An analysis of available data sets is followed by the development of STF models based on different types of input data (pH, DO and temperature). The choice of model for reconstruction is based on the comparison of the results with monthly observations of algae concentrations at Schnackenburg during the years 1985—2001 and daily observations at Cumlosen in the years 1997—1999. The stochastic transfer function model based on DO concentrations and temperature gives the best fit to the data, followed closely by the model additionally using pH values as an input.