Thedeploymentofenvironmentalsensorshasgeneratedaninterestinreal-timeapplicationsofthedatatheycollect.Thisresearchdevelopsareal-timeanomalydetectionmethodforenvironmentaldatastreamsthatcanbeusedtoidentifydatathatdeviatefromhistoricalpatterns.Themethodisbasedonanautoregressivedata-driven