Variabilities and uncertainties in important coastal water variables
The focus of this work is on general patterns in uncertainty as well as temporal and spatial variability in key water variables in coastal science and management. These patterns are essential since they regulate how many samples must be taken to get reliable mean or median values characterising coastal water quality and which variables are most suitable for monitoring and predictive modelling. We present results concerning coefficients of variation, correlations, regressions, variations in data from different time periods, and confidence intervals for empirical mean values. We use data from Ringkøbing Fjord (Denmark, N. Europe), Chesapeake Bay (Eastern U.S.) and other coastal marine sites to illustrate the basic principles related to patterns in variability. We have shown that total and particulate N and P generally have much lower coefficients of variability (CV) than dissolved inorganic nutrient fractions. The latter are, hence, of limited use in predictive models for coastal management. Total nitrogen (N) and phosphorus (P) were, on the other hand, found to be useful predictors of two standard bioindicators, the Secchi depth (a measure of water clarity) and chlorophyll-a concentrations (a measure of phytoplankton biomass or production).