Time series analysis of water quality data from Lake Ontario: implications for the measurement of water quality in large and small lakes

SUMMARY 1. The normal strategy of monitoring water quality is to sample such parameters as chlorophyll no more than weekly. A preferable strategy is to first define the natural periodicities in the water body and then to set up a sampling scheme that takes into account the natural scales of variance in physical, chemical and biological parameters. Failure to do so leads to aliased and biased estimates of means and variances and an inability to interpret the underlying physical and biological mechanisms. 2. The natural scales of variance vary with basin size. In lakes, physical and biological processes overlap at scales of from 1 to 15 days. Time series analysis of daily data from Lake Ontario and other lakes showed how the means and variances of the data sets were determined by the physical and biological processes in the water columns and displayed the fundamental lags in the systems. Even in small lakes and reservoirs, advective processes were of great importance. Advection became the dominant process in Lake Ontario. Time lags and advection made simple correlations of physical and biological parameters meaningless. 3. Decimation of the daily data sets revealed the statistical dangers of less frequent sampling. The desirable frequency of sampling was shown to be a function of the physics of the mixed layer, the turnover times of the nutrient pools, and the biological activity. Data from the three lakes graphically demonstrated the inadequacy of normal sampling frequencies.

[1]  F. M. Boyce.  Response of the Coastal Boundary Layer on the North Shore of Lake Ontario to a Fall Storm , 1977 .

[2]  G. Csanady.  Response of Large Stratified Lakes to Wind , 1972 .

[3]  K. Emery,et al.  Surface circulation of lakes and nearly land-locked seas. , 1973, Proceedings of the National Academy of Sciences of the United States of America.

[4]  G. T. Csanady,et al.  On the Equilibrium Shape of the Thermocline in a Shore Zone , 1971 .

[5]  J. Blanton.  Some Characteristics of Nearshore Currents along the North Shore of Lake Ontario , 1974 .

[6]  E. Lindstrom,et al.  A Simple Model of Lake Ontario's Coastal Boundary Layer , 1977 .

[7]  G. Harris,et al.  Phytoplankton population dynamics of a small reservoir: effect of intermittent mixing on phytoplankton succession and the growth of blue-green algae , 1984 .

[8]  F. M. Boyce.  Some Aspects of Great Lakes Physics of Importance to Biological and Chemical Processes , 1974 .

[9]  A. M. Trimbee,et al.  Use of time-series analysis to demonstrate advection rates of different variables in a small lake , 1983 .

[10]  T. Platt,et al.  Day‐to‐day variations in the spring‐summer photosynthetic parameters of coastal marine phytoplankton , 1983 .

[11]  G. Harris,et al.  Phytoplankton population dynamics of a small reservoir: physical/biological coupling and the time scales of community change , 1986 .

[12]  H. Hansen.  The significance of discrete baltic nutrient samples discussed on the basis of continuous measurements , 1984 .

[13]  R. Vollenweider,et al.  Seasonal Variations of Temperature and Nutrients in the Surface Waters of Lakes Ontario and Erie , 1974 .

[14]  A. M. Trimbee,et al.  Phytoplankton population dynamics of a small reservoir: use of sedimentation traps to quantify the loss of diatoms and recruitment of summer bloom-forming blue-green algae , 1984 .

[15]  J. Bennett.  On the Dynamics of Wind-Driven Lake Currents , 1974 .

[16]  G. Harris,et al.  Physical variability and phytoplankton communities: VI. Day to day changes in primary productivity and species abundance , 1984, Archiv für Hydrobiologie.

[17]  Jeffrey L. Star,et al.  Spectral-Analysis - a Caveat , 1981 .

[18]  Colin S. Reynolds,et al.  Growth- and loss-rate responses of phytoplankton to intermittent artificial mixing and their potential application to the control of planktonic algal biomass , 1984 .