> the error margin for the estimate [of number of trees on Earth] ranged between one trillion and ten trillion A moment's googling gives primary forest being 1

> the error margin for the estimate [of number of trees on Earth] ranged between one trillion and ten trillion A moment's googling gives primary forest being 1/3 of Earth land area and 40M km^2 ("world forest square km"), and primary forest tree density being 50k to 100k trees per km^2 ("trees per square km"). So there are at least 1T trees, as forests alone have more. And exceeding 10T would require non-forests to average at least half the tree density of forest, which seems unlikely. Suggesting bounds of 1T and 10T trees. Just a reminder that rough quantitative reasoning and Fermi problem solving are powerful. Especially when approached as an exercise in order-of-magnitude bounding, rather than point estimate. "How many trees are there in the world, is it more like 1, 10, 100, 1k, etc? Can anyone suggest a low bound?" "There's a tree outside the window. So one tree." "How confident are we? Should we consider that a hard or soft bound?" ... "Ok, a hard lower bound of 1 tree." "Can anyone suggest an upper bound?" "Ok, sigh, Jim?" "There can't be more trees on Earth than atoms in the visible Universe, because all Earth trees are part of the Universe, and each is made of lots of atoms! So a hard upper bound of 10^80 trees!" "Ok, is everyone ok with a hard upper bound of 10^80?" ... "Can anyone suggest some narrower bound?" ... The phrase "I've no idea how many/much/etc" seems said far more often then it's true. You may not know it to some needed accuracy. But even young kids can be taught to estimate bounds. Which often turn out quite narrow enough to move on with.