Serious question: A paper in GRL last year found, using satellites, that lakes comprise about 3.7% of the nonglaciated land area on the planet.
They defined a "lake" as more than 2000 m2 in size -- about half a football field. They count 117 million of them, with a total size is 5 million km2 -- half the size of the U.S.
So the average lake size is 42,000 m2 (207 m)2.
Are the temperature trends of lake surfaces being monitored, as we do on land? If not, isn't that a potential bias in the long-term record? (I don't know which way it would go.) Or is it that the exclusion of lakes lead to a small enough error?