Optimising data quality from environmental monitoring stations

Dave Johnstone - NZ NIWA , 28 November, 2004

A wide variety of electronic sensors are now used to continuously record environmental data. Sensor manufacturers specify the quality of data that can be expected from their instruments, provided calibration is maintained. However a sensor is usually one of a chain of hardware and software components comprising a measurement / recording system, each of which can contribute to uncertainty in the recorded data. The uncertainty in the recorded data attributable to the sensor performance may be a minor part of the overall uncertainty. An understanding of the potential for degradation of data at each of the components / stages allows the system to be designed to minimise degradation and enable confident interpretation of the data. The monitoring systems considered in this paper comprise discrete sensors and a data logger connected by cables, with examples from hydrometric, meteorological, and water quality stations. The elements of a system that contribute to uncertainty in recorded data are identified, and methods for minimising the uncertainty are described.