A technique for using satellite radar altimetry data to estimate the t
emporal variation of the water level in moderate to large lakes and en
closed seas is described. Great Lakes data from the first 2 years of t
he U.S. Navy's Geosat Exact Repeat Mission (November 1986 to November
1988), for which there is an improved orbit, are used to demonstrate t
he technique. The Geosat results are compared to the lake level data c
ollected by the Great Lakes Section, National Ocean Service, National
Oceanic and Atmospheric Administration, and are found to reproduce the
temporal variations of the five major lakes with root-mean-square err
or (rms) ranging from 9.4 to 13.8 cm and a combined average of 11.1 cm
. Geosat data are also analyzed for Lake St. Clair, representing a mod
erate-sized lake, with a resulting rms of 17.0 cm. During this study p
eriod, the water level in the Great Lakes varied in a typical annual c
ycle of about 0.2 m (0.5 m for Lake Ontario) superimposed on a general
decline of approximately 0.5 m. The altimeter data reproduced the gen
eral decline reasonably well for the lakes, but the annual cycle was o
bscured in some lakes due to systematic errors in the altimeter data.
Current and future altimetry missions will have markedly improved accu
racy which will permit many moderate (25 km diameter) or larger lakes
or enclosed seas to be routinely monitored.