(1) above detection levels, at levels that are
(2) practically achievable, and
(3) where harm is likely. Concern over barely detectable levels incapable of being measured is for unrealistic perfectionists.
(1) the level of significance is 95%, which is generally considered the bare minimum level of confidence in hard sciences;
(2) there are many mechanisms that could have generated the trend including increasing solar intensity, decreasing gamma-ray flux, urban heat influence, aerosols, and internal climate cycles; and
(3) the temperature trend had a break or sudden increase in 1976, noted by many other researchers, that argues against a harmful long-term trend in temperature for reasons offered in a paper by Stockwell.