Safecast was started because of a lack of what we considered to be useful data. We’ve talked before about the difference between correct and useful information and this is a perfect example. We saw data published with no information about how readings were taken, or where. We saw data published that was questionable, but without any way to trace the source. We even felt that “correct” or “valid” data, such as a single radiation data point being given as an average of an entire city, wasn’t actually useful and possibly did more harm in the way of public understanding, and rather than just complain about it we set out to provide a better option. Since that time a number of people and groups have started taking readings and publishing them – and we’re glad to see more data available to the public. That said, not all data is created equal, so we wanted to publish a few best practices that we consider paramount – We won’t allow data that doesn’t match up to these standards into the Safecast database, and we encourage other groups to adopt these standards as well.
- Every data point must have time stamp and GPS coordinates. City averages are unacceptable.
- Device and user must have unique ID to provide ability to identify and track errors (user can remain anonymous but still needs ID)
- The device used must have publicly available specs, including sensor size and manufacturer.
** We prefer and recommend the 2″ pancake produced by LND.
- If data is submitted to Safecast, data must be formatted correctly with all applicable metadata.
- Measurements should be taken at approx 1m, with care to avoid interference etc.
- Measurements should be in CPM, and note what kind of shielding (if any) was used.
We feel that data integrity is the most important aspect of what we do – if our data can’t be trusted then everything else we do is pointless, and hope these standards we hold ourselves to and this information is helpful to others as well.