Safecast – a Citizen Science initiative for ambient dose rate mapping; Quality assurance issues.

Safecast has been initiated in 2011 in Japan as response to the perceived inadequacy of official information policy about radioactive contamination. It is based on measurements of ambient dose rate (ADR) by numerous volunteers using a standardized monitor, called SAFECAST bGeigie Nano. In essence, it consists of a Geiger counter and a GPS module, data (ADR, GPS coordinates, date/time) are recorded on an SD card if operated in its survey mode.

The project quickly expanded world-wide and by end 2020, over 150 million measurements were recorded, however by far not uniformly distributed over the world (https://map.safecast.org/ ). Evidently, such amount of data cannot be reasonably acquired by institutional surveying. On the other hand, professionals can be expected to follow metrological quality assurance (QA) standards, which is usually not the case for members of the public who are mostly laypeople in metrology.

Thus, impressive as the Safecast map is, it raises questions related to QA. This is relevant for interpretation of the ADR values shown on the map, and their uncertainty and resulting reliability. We propose to distinguish between two aspects of metrological QA regarding monitoring in the context of citizen science.

(1) Metrology proper, which pertains to characterization of the measurement procedure, from sampling protocols to physical behaviour of the instrument and resulting uncertainty; this is of course equally true also for professional measuring.

(2) Real-world handling: not being familiar with metrological QA concepts, in general, it can be expected that citizen scientists deviate from QA standards more frequently and more severely than professionals. This adds to the uncertainty budget of reported values. Uncertainty impairs interpretability.

In this contribution, we report current metrological knowledge of the bGeigie Nano in the sense of aspect (1). Further, we discuss how QA in the sense of aspect (2) can be approached. We report experiments of repeated realistic handling, i.e. without caring for particularly controlled laboratory or well-defined field conditions (as in (1)) and of intentional mishandling.

It appears that QA type (2) is the more serious issue, both by contribution to the uncertainty budget and by difficulty in handling it. While the Safecast map provides – in some regions - an astonishing dense database, one must be cautious about interpreting local data, if the measurement circumstances are not known, which is the usual case. One element of addressing the problem consists in instruction of participants about correct usage.

In response to certain technical issues of the bGeigie Nano which derogate its performance, SÚRO developed an alternative but conceptually similar device called CzechRad (details in https://github.com/juhele/CzechRad) whose metrological characterization is ongoing.