Methods for evaluation at biological field trials

A field trial was performed within TRIBALS (FOI, TNO and DRDC Suffield) during one week, 14th to 18th September 2009, in Umeå. The objective of the trial was to compare the results from the Canadian and Swedish reference system, and to evaluate the previously used approach of determining limit of detection (LOD) at an outdoor field trial. The theory used to obtain limit of detection for detectors of bioaerosols from laboratory test and evaluation can also be used for sensor data from field trials. The theory is based on standard deviation of the line obtained by plotting detector signal versus reference signal. The higher variability in field trial data results in larger limit of detections. The suggestion in this report is to interpret LOD obtained from field trials as the lowest concentration one can get reliable alarms at. This field trial gives a value of about 24 ACPLA, with a 1% risk of false positive alarms. The comparison between reference STA sampler systems revealed, within measurement uncertainty, that the Swedish and Canadian systems recorded the same ACPLA concentrations.