Topographic and sedimentological evidence indicates that stagnant ice
conditions prevailed during retreat of many Pleistocene valley glacier
s in the San Juan Mountains, Colorado. I use topographic data from man
y of these valleys to test two conceptual models that predict the most
likely position for a stagnant ice margin to develop during valley gl
acier retreat. In the first model, valley hypsometry controls the rate
of accumulation area loss for a rise in equilibrium line altitude (EL
A). The faster accumulation area is lost, the faster the glacier termi
nus must retreat, increasing the likelihood of ice stagnation. In the
second model, a stagnant margin will develop if a topographic obstacle
thins the glacier to a critical thickness, retarding internal deforma
tion and pinching off a section of un-nourished ice. Comparison of mod
eling results with field evidence indicates that topographic obstacles
do not force the development of stagnant ice margins in the San Juans
. Instead, valley hypsometry, in particular the valley slope at the pa
leo-ELA, appears to be the primary control of ice stagnation. For a co
nstant ELA rise rate, gentle valley slopes force ice stagnation (by in
creasing retreat rate) whereas steep slopes encourage active retreat.
Ice stagnation is prevalent in the San Juan Volcanic Field because inc
ompetent volcanic rocks are easily eroded to produce low valley gradie
nts. This finding that the slope at the ELA is an important control on
the development of a stagnant margin is supported by the abundance of
stagnant ice deposits in continental settings where the slope at the
ELA is very low.