Water Resources Research · 2014 · 67 citations · 65 references
Life SciencesEcologyEnvironmental SciencesBayesian Data AnalysisEnvironmental ModelingFreshwater EcologyAquatic Microbial EcologyBayesian Model BuildingEcological ModelingHarmful Algal BloomsLake EcologyBayesian Hurdle ModelsBloom ForecastingBloom Prediction ModelsBloom-forming CyanobacteriaEnvironmental FactorsResidence TimeSpecies AbundanceToxic Algal Blooms
Abstract There have been increasing reports of harmful algal blooms (HABs) worldwide. However, the factors that influence cyanobacteria dominance and HAB formation can be site‐specific and idiosyncratic, making prediction challenging. The drivers of cyanobacteria blooms in Lake Paldang, South Korea, the summer climate of which is strongly affected by the East Asian monsoon, may differ from those in well‐studied North American lakes. Using the observational data sampled during the growing season in 2007–2011, a Bayesian hurdle Poisson model was developed to predict cyanobacteria abundance in the lake. The model allowed cyanobacteria absence (zero count) and nonzero cyanobacteria counts to be modeled as functions of different environmental factors. The model predictions demonstrated that the principal factor that determines the success of cyanobacteria was temperature. Combined with high temperature, increased residence time indicated by low outflow rates appeared to increase the probability of cyanobacteria occurrence. A stable water column, represented by low suspended solids, and high temperature were the requirements for high abundance of cyanobacteria. Our model results had management implications; the model can be used to forecast cyanobacteria watch or alert levels probabilistically and develop mitigation strategies of cyanobacteria blooms.
65
R: A Language and Environment for Statistical Computing
R Core Team · 2000 · 353K citations · Full text
Natural Sciences, Computer Science, Statistical Computing +10
Hans W. Paerl, Jef Huisman · Science · 2008 · 2.8K citations