Nordgaard, AndersGnauck, AlbrechtHeinrich, Ralph2019-09-162019-09-162003https://dl.gi.de/handle/20.500.12116/27106In this paper we assess the power of the Hirsch & Slack test for trend in water quality data from a number of small rivers in Sweden. In particular, we compare estimated power functions from tests applied to observations taken monthly and quarterly. The power function is estimated through a simulation study with data generated by bootstrap resampling in contrast with common methods based on the normal distribution. The results show a substantial loss in power when the sampling frequency is lowered from monthly to quarterly data. The loss depends on both the river and the variable analysed and this information can be transferred to optimisation procedures in the design of monitoring programs.Text/Conference Paper