Forecasting Phaeocystis globosa blooms in the Dutch coast by an integrated numerical and decision tree model
Titel:
Forecasting Phaeocystis globosa blooms in the Dutch coast by an integrated numerical and decision tree model
Auteur:
Chen, Qiuwen Mynett, Arthur E.
Verschenen in:
Aquatic ecosystem health & management
Paginering:
Jaargang 9 (2006) nr. 3 pagina's 357-364
Jaar:
2006-09-01
Inhoud:
In Dutch coastal waters, in particular the Noordwijk transect and the Wadden Sea, spring algal blooms have occurred almost every year which are usually identified as diatoms or Dinophysis, and followed by Phaeocystis globosa (P. globosa). The blooms of P. globosa incurred adverse impacts on shellfish farming because of the resulted anoxic conditions, and on recreation due to odour and foam. Funded by EU FP5 Harmful Algal Blooms Expert System (HABES) project, the objective of this research was to systematically assimilate the various available knowledge and limited data to develop innovative and practical methods to forecast P. globosa blooms. From previous studies in the Dutch coast, water column irradiance was identified to be a triggering factor to the blooms of P. globosa. Phosphate was also confirmed to be an important limiting factor. The available water column irradiance was related to the vertical position of the species. In this paper, a one-dimensional vertical advection-diffusion model was used to investigate the effects of sinking (buoyancy) and turbulence mixing on the vertical profile of P. globosa concentrations. The numerical model was then integrated with decision tree method to predict possible P. globosa blooms by using the available chemical and biological data. The results of the case study at station Noordwijk 70 were shown to be comparable to the field observations. The research demonstrated that for algal blooms forecasting, an integrated approach gave a good perspective and deserves further efforts because (1) the numerical module could provide understanding of the fundamental mechanisms; (2) the limited water quality and biological data, and the empirical knowledge from experts could be incorporated into the rule-based model.
Tijdschrift beschrijving