Small dataloggers with integrated sensors can be very useful as a less-expensive means of obtaining replication and spatially distributed measurements. Though they often rely heavily on the data, researchers rarely assess the precision and accuracy of equipment before purchase or before deployment.
To assess accuracy and precision as well as demonstrate a method, five each of three types of low-cost temperature loggers (UA-001-64, DS1921G, DS1922L) were placed in an environmental chamber along with calibrated references (CR6, 109 thermisters). Temperature was varied from -20 to 50 C at 5 degree increments and held at each increment for 60 minutes to ensure enough time for equilibration.
Results/Conclusions
All loggers showed good agreement with the reference as well as good precision during times when temperatures were stable; generally, within 0.5 degrees C. However, not surprisingly, the accuracy and precision were lower when temperature was changing rapidly with differences of 1 to 2 degrees C.
Therefore, under fairly ideal conditions, these loggers should perform well. However, particularly when measuring air temperature, there can be a great deal of error associated with the installation of any sensor. Errors of several degrees C have been shown as a result of inadequate or lack of shielding from solar radiation.
A key recommendation when using any instrumentation is to verify all the sensors agree with each other under the same conditions and to use this information both in experimental planning and in subsequent analyses. Preferably, these calibration checks should be performed both before and after each deployment, but can be as simple as one to three points that represent the range of interest. At a minimum, the specifications published by manufacturers should explicitly be taken into consideration during planning, analyses, and publication.