They are based on models developed from multiple measurements from a variety of sensors, and the denialists are always questioning their assumptions, accuracy and consistency, not to mention the leftist political motives of the climatologists and government agencies that publish the data.

The ice extent data is measured optically with IR radiometers and the sensors see through clouds, so they work day and night regardless of the weather.

The data still has to be calibrated, but what matters is not the actual readings, but how those readings vary over time. The dramatic drop in summer Arctic sea ice extent over the last 37 years
(about 40%), even when superimposed on the noise due to random variability and meteorological phenomena, is difficult to explain away with conspiracy theories.

Over the next few days, I will be publishing the August 1979-2016
graph, and early in October the successor to the above graphic, the September 1979-2016 figures from NSIDC. I promise there will be no more polar updates from me after that. The denialists have had every opportunity to present data in rebuttal and they have been conspicuous by their absence.

Still this new thickness data may come in handy in the future. I expect it can be used as evidence in the trials.