Page 6 - i1052-5173-27-8

from 19.3 km2 to 1.72 km2, representing a
91% loss (Kincaid, 2007). From 2000 to
2002 alone, surface area decreased from
2.326 km2 to 2.152 km2, or by 7.48%
(Klein and Kincaid, 2006). The rate of
retreat accelerated from 1988 to 2005, even
while precipitation (partly as rain) actually
increased (Kincaid, 2007). When ice con-
tracts in area and thickness, the ice within
the glacier can also be affected by melting.
Snow pits and cores at the Quelccaya ice
cap in southern Peru reveal that since the
late 1970s the seasonal oxygen isotopic
(d18O) variations have been homogenized
by meltwater percolating through the top
20 to 30 m of firn. This homogenization
compromises the long-term seasonally
resolved record of past climate variations.
This finding is consistent with analyses of
shallow cores throughout the Cordillera
Blanca of northern Peru (Davis et al.,
1995). Radiocarbon dates from wetland
plants exposed by the retreating margins of
Quelccaya ice demonstrate that, for
>~6,300 years, this ice cap has not been
smaller than it is today (Thompson et al.,
2013). Rapid retreat of the ice margin con-
tinues to expose such evidence.
Photography
Our collaboration features the work of
the Extreme Ice Survey (EIS), a non-
governmental organization founded to
photograph the retreat of glaciers. Photo
couplets of ice retreat have been both
coincidentally and intentionally collected.
Ernest Shackleton’s expedition on HMS
Endurance, for example, collected historic
photos of the extent of ice on South
Georgia that can be compared to modern
photos (see https://vimeo.com/125634374).
Figure 2 presents photo couplets of glacial
retreat in Alaska, Iceland, Switzerland, and
Peru, where lateral retreat and thinning are
apparent. In a similar vein, the EIS has
amassed >1.1 million images recorded by
43 cameras observing 24 glaciers in
Alaska, the Rockies, the Andes, South
Georgia, Antarctica, the Alps, Iceland, and
the Himalayas. Ice: Portraits of Vanishing
Glaciers (Balog, 2012) provides graphic
evidence in print form. Such time-lapse
imagery has also been assembled into
videos that display ice retreat, such as
“Extreme Ice” (https://www.youtube.com/
watch?v=6scs-Q-Ut_E). The film Chasing
Ice is critically acclaimed for its portrayal a rate of 0.16 °C annually and 0.32 °C per
of glacial retreat. The film captured spell-
binding imagery of perhaps the largest
6 GSA Today | August 2017
calving event ever witnessed, when
Greenland’s Ilulissat glacier discharged
a section of its terminus that measured
~5 km wide, 1 km thick, and 1.5 km long.
CAUSES OF ICE LOSS
Extensive literature shows that the
ongoing loss of mass from glaciers is
being caused primarily by warming over
those glaciers and that this warming is, in
turn, being caused primarily by the rising
CO2 concentrations in the atmosphere.
The reports of the Intergovernmental
Panel on Climate Change (IPCC) provide
useful starting points for understanding
the linkage between temperature and the
mass balance of a glacier (e.g., Lemke et
al., 2007), and IPCC (2013) examines the
history and causes of warming. Glaciers
can respond to changes in accumulation
of snowfall, seasonality of temperature,
cloudiness, and other factors. The advance
or retreat of a single glacier may be dif-
ficult to interpret without targeted stud-
ies, but literature summarized in these
sources shows that for a large suite of gla- melting (Bradley et al., 2006).
ciers ending on land, retreat is primarily
driven by atmospheric warming.
Paleoclimate information contributes in
fundamental ways to the strong evidence
that warming temperature is the primary
driver of the recent acceleration of ice
retreat (Thompson et al., 2011, 2013). For
example, evidence for warming is associ-
ated with ice retreat in the tropical Andes
(Rabatel et al., 2013). Arendt et al. (2013) much smaller than their polar counterparts.
found that mean summer temperatures
derived from ground and lower tropo-
sphere records were good predictors of
GRACE-derived summer mass balances
in Gulf of Alaska glaciers, capturing 59%
and 72% of the variability. In the context
of the ice retreat in New Guinea from
1972 to 1987, mean monthly atmospheric
temperature was the only climate variable
that changed in a statistically significant
way (+0.24 °C; Klein and Kincaid, 2006).
Warming is also seen throughout the
Tibetan Plateau (now sometimes charac-
terized as the Third Pole [TP]), where
meteorological data show that surface
temperatures are rising faster at higher
elevations than at lower elevations (Liu
and Chen, 2000). On average, the tem-
perature on the TP has been increasing at
decade during winter.
On decadal and longer time scales,
climate models project that greenhouse-
gas–forced warming will drive tempera-
tures to rise faster with elevation, with this
vertical amplification being greatest in the
tropics due to feedbacks involving upper-
tropospheric humidity, as well as snow
albedo and surface-based and water-vapor
feedbacks (Ramaswamy et al., 2006;
Randall et al., 2007; Pepin et al., 2015).
Results from general circulation models
indicate that the combined water-vapor/
lapse-rate feedback provides the largest
positive radiative reaction and that this
effect alone roughly doubles the warming
in response to forcing by greenhouse
gases. As a result, the projected changes in
mean annual free-air temperatures show
twice as much warming at higher eleva-
tions in the tropics as is predicted at
Earth’s surface generally (Bradley et al.,
2006). These projections are consistent
with the recently documented rise of the
free-air 0 °C isotherm in the tropical atmo-
sphere (Bradley et al., 2009). Furthermore,
as more dark land surface is exposed,
absorption of the intense higher-elevation
radiation increases, thus accelerating the
Low- to mid-latitude glaciers are
extremely susceptible to such warming. In
accord with model predictions of warming,
high-elevation tropical glaciers appear to
be responding with an accelerating rate
of glacier loss (Coudrain et al., 2005;
Thompson et al., 2006). Smaller glaciers
respond more rapidly to climate changes,
and these mountain glaciers are generally
These ice masses are also particularly sen-
sitive to small changes in ambient tempera-
tures, because they already exist very close
to the melting point.
We again emphasize that many environ-
mental factors affect glaciers, and one
glacier may change for many reasons. As
glaciers shrink, the insulating effect of a
debris cover that slows further melting
may become more important, joining other
factors influencing glaciers, some of which
are mentioned above. Thus, proper charac-
terization of ongoing trends requires moni-
toring of many glaciers in many places,
together with targeted studies of selected
glaciers to better characterize controls.
Taken together, though, the full scholar-
ship as summarized above gives high con-
fidence that warming caused primarily
by human release of greenhouse gases is
causing the retreat of glaciers.