Abstract
Assessing environmental risk is useful for preventing adverse effects on human health in highly polluted cities. We design a criterion for environmental monitoring based on an attribute control chart for the number of dangerous contaminant levels when the concentration to be monitored follows a Birnbaum-Saunders distribution. This distribution is being widely applied to environmental data. We provide a novel justification for its usage in environmental sciences. The control coefficient and the minimum inspection concentration for the designed criterion are determined to yield the specified in-control average run length, whereas the out-of-control case is obtained according to a shift in the target mean. A simulation study is conducted to evaluate the proposed criterion, which reports its performance to provide earlier alerts of out-of-control processes. An application with real-world environmental data is carried out to validate its coherence with what is reported by the health authority.
Original language | English |
---|---|
Pages (from-to) | 463-476 |
Number of pages | 14 |
Journal | Environmetrics |
Volume | 26 |
Issue number | 7 |
DOIs | |
State | Published - Nov 2015 |
Externally published | Yes |
Keywords
- Average run length
- Inspection point
- Moment estimation
- Monte Carlo simulation
- Np control chart
- R software