Environmental concentrations as ratios of random variables

Abstract Human-induced environmental change increasingly threatens the stability of socio-ecological systems. Careful statistical characterization concentrations is critical to quantify and predict consequences such changes on human ecosystems conditions. However, while are naturally defined as ratio between solute mass solvent volume, they have rarely been treated such, typically limiting analysis familiar distributions generically used for any other variable. To address this gap, we propose a more general framework that leverages their definition explicitly ratios random variables. We show resulting models accurately describe behavior nitrate plus nitrite in US rivers salt concentration estuaries Everglades by accounting heavy tails potentially emerging when water volume fluctuates around low values. Models preclude presence related high probability extreme could significantly undermine accuracy diagnostic frameworks effectiveness mitigation interventions, especially soil contamination characterized (i.e. moisture) frequently approaching zero.