Estimates of Regional Equilibrium Line Altitudes and Net Mass Balance from MODIS Imagery

Glacier mass balance is a key variable used to assess the health of glaciers and ice sheets. Estimates of glacier mass balance are required to model the dynamic response of glaciers and ice sheets to climate change, estimate sea-level contribution from surface melt, and document the response of glaciers to climate forcing. Annually resolved estimates of regional mass balance for mountain ranges is often inferred from a sparse network of ground-based measurements of mass balance for individual glaciers. Given that net mass balance is highly correlated with the annual equilibrium line altitude (ELA), we develop an automated approach to estimate the ELA, and by inference net mass balance, on large glaciers and icefields using MODIS 250 m imagery (MOD02QKM). We discriminate areas of bare ice and snow/firn using the product of MODIS' red (0.620 - 0.670 μ m) and near infrared (0.841 - 0.876 μ m) bands. To assess the skill in estimating glacier ELAs, we compare ELAs derived from (1) manual delineation and (2) unsupervised classification of the band product to ground-based observations of ELA and net mass balance at seven long term mass-balance monitoring sites in western North America (Gulkana, Wolverine, Lemon Creek, Taku, Place, Peyto, and South Cascade). Spatial and temporal variations in MODIS-derived ELAs provide an opportunity to validate regional mass-balance models, estimate surface melt contributions to sea-level rise, and examine the cryospheric response to climate change.