LAUSR

Under the EU Water Framework Directive, the status of a surface water body (wb) is determined from a water sampling program. The reliability of this determination partly depends on the frequency of such sampling and is the subject of this paper. Available water quality data were acquired from the national environmental monitoring system in Greater Poland (Wielkopolska). At any given monitoring station, an assumption of normally distributed data was checked for all water quality indicators (wqi) and relevant statistical parameters identifi ed. For particular sampling frequencies, a Monte Carlo simulation model was used to generate synthetic wqi data series from which wb status could be estimated. Assuming that 24 annual measurements of all wqi is the maximum economically affordable sampling frequency and taking this frequency as a reference, it has been shown that in about 22% of cases, a water quality class assessed from sampling at the standard frequency of 12 times per year is inaccurate. In less than 50% of cases was the reference assessment better than classifi cations based on lower frequencies. Nevertheless, in 33% of cases the indicator class was correctly assessed from measurements taken 4 times annually. The correlation between water quality class assessment and sampling frequency is not simple nor can it be arbitrarily assumed. When choosing the sampling frequency as a basis for reliable water quality assessment it is necessary to take into account characteristics of natural and man-made pressures acting on rivers in their catchments.