Some radio wags have contended that it would be cheaper to fiddle with the Earth's actual rotation, than to keep screwing around with leap seconds. How much effort would it really require to speed up or slow down the Earth's rotation by one second in one year?
So let's do the calculation. The angular velocity is to have a relative increase of 1 s/1 yr = 3e-8. Assume the Earth has uniform density, radius R=6.4e6 m, mass M=6.0e24 kg. We will transfer a shell of mass m and thickness d uniform over the Earth's surface to or from the poles. The moment of inertia must have a relative decrease of 3e-8. For the Earth, I=(2/5)MRR, for the shell i=(2/3)mRR. Then i/I=(5/3)(m/M)=3e-8, or m/M=2e-8. This is the same as the volume ratio, v/V=[(4pi)RRd]/[(4pi/3)RRR]=3(d/R)=2e-8, and the depth is d=(2/3)e-8 R = 0.042 m. If we use the ocean for this purpose, then about 70% of the surface is available and the average density of the Earth is 5.5 times that of water. To speed up the Earth and eliminate the leap second, we need to reduce sea level by 0.33 m, about 1 foot, and transfer that water to the South polar ice cap (so it is supported by land). This is easily in the range of an ice age. On the other hand, it's not clear that humans have any influence over global temperatures, for, Al Gore fanboys and Kyoto treaty signers notwithstanding, Earth's climate is controlled by the Sun. Gene __________________________________________ Yahoo! DSL Â Something to write home about. Just $16.99/mo. or less. dsl.yahoo.com