The environment may be altered by marine renewable energy developments, which include offshore wind turbines, surface wave converters, and tidal turbines. To measure their impact, it is crucial to first study pre-development conditions, but indicators tracking these conditions may include variability that can be above and beyond the ability of standard models to characterize. In a new paper, the performance of 13 different types of models is tested, with three particular methods performing well under different conditions: vector regression, random forests, and state-space models. The findings can be applied to measure baseline conditions for any human intervention in a natural system, not just to the monitoring of new marine renewable energy developments. The research was conducted by SAFS MS student Hannah Linder and Prof. John Horne, together with Eric Ward from the Northwest Fisheries Science Center, and appears in the journal Ecological Indicators.