Using reanalysis data to characterize Arctic and Sub-Arctic glaciers

by Ourada, Quin

The net mass balance fluctuations of Arctic and Sub-Arctic glaciers, north of 47.5°

North latitude, are described over a 45-year period from 1957 to 2002 using two

parameters derived from a gridded climatology reanalysis. Variability among 185

measured glaciers was represented according to two main components. The first

component represents the temporal variability of the net balance series derived from a

statistically defined geographic region. The second component represents the glacier-specific

variability in the net balance series (the amplitude of variability). Each

component was numerically derived using gridded monthly mean temperature and

precipitation data at a 2.5° spatial resolution. These two main components of the net

balance series can be determined from simple glacier location data. The temporal

variability component was determined using patterns of warm season temperature that

were then used to define the statistically correlated regions. The glacier-specific

component was approximated along continuum of continentality. Continentality was

based on the relative ratio of warm to cool season temperature at the glacier location,

which was then normalized and differenced using precipitation magnitude.

Ultimately 21 distinct geographic regions containing at least one representative glacier

were defined for the first, temporal component. In data-rich regions, such as the Alps

and Scandinavia, spatial variability was identified on a finer scale than individual

mountain ranges. The temporal evolution of measured net balance series within each

region were more closely related to each other than measured mass balance on the

scale of mountain ranges. This temporal signal can be considered the most likely

temporal signal that would be characteristic of unmeasured glaciers within the spatial

extent of the region. This spatial extent defined for each region is specific, and based

on similarities in physical climatology as opposed to more vaguely defined regions

based on mountain ranges or other geographic features. The, second, glacier specific

component of the series was related to net balance standard deviations (58% variance

explained), balance amplitude (55% variance explained) and climate sensitivity (56%

variance of temperature sensitivity and 52% variance of precipitation sensitivity

explained) for measured glaciers. The normalization process resulted in a glacier

continuum ranging from -1 to 1 to describe the relative position of a glacier along a

continuum or wet-maritime to dry continental.

These two pieces of information can be used together to approximate a large

component of the net balance series for an unmeasured glacier based on location

alone. Representing unmeasured glaciers in this manner is, by no means, a substitute

for actual field measurements or complex and highly parameterized mass balance

models. This approach is also limited in accuracy by the spatial resolution of the

gridded climatologies used, which at this time are still quite coarse, 2.5°. However, in

lieu of more detailed data, a simple approximation of glacier mass balance can be

made and those measured glaciers most likely to exhibit similar characteristics can be

identified to assist in tuning parameters for a more detailed mass balance model.