Abstract
Abstract We investigate relationships between synoptic-scale atmospheric variability and the mass-balance of 13 Andean glaciers (located 16–55° S) using Pearson correlation coefficients (PCCs) and multiple regressions. We then train empirical glacier mass-balance models (EGMs) in a cross-validated multiple regression procedure for each glacier. We find four distinct glaciological zones with regard to their climatic controls: (1) The mass-balance of the Outer Tropics glaciers is linked to temperature and the El Niño-Southern Oscillation (PCC ⩽ 0.6), (2) glaciers of the Desert Andes are mainly controlled by zonal wind intensity (PCC ⩽ 0.9) and the Antarctic Oscillation (PCC ⩽0.6), (3) the mass-balance of the Central Andes glaciers is primarily correlated with precipitation anomalies (PCC ⩽ 0.8), and (4) the glacier of the Fuegian Andes is controlled by winter precipitation (PCC ≈ 0.7) and summer temperature (PCC ≈ −0.9). Mass-balance data in the Lakes District and Patagonian Andes zones, where most glaciers are located, are too sparse for a robust detection of synoptic-scale climatic controls. The EGMs yield R2 values of ~ 0.45 on average and ⩽ 0.74 for the glaciers of the Desert Andes. The EGMs presented here do not consider glacier dynamics or geometry and are therefore only suitable for short-term predictions.
Highlights
A large fraction of southern hemisphere glaciers is located in the Andes of South America (Pfeffer and others, 2014)
The first group consists of the Outer Tropics glaciers in Boliva (Zongo (ZON), Charquini Sur (CHS), Chacaltaya (CHT)) which are all significantly negatively correlated to summer 2 m air temperature (t) (α ≤ 0.05)
The Ba of Echaurren Norte (ECH) is significantly positively correlated with the Multivariate El Niño-Southern Oscillation (ENSO) Index, Antarctic Oscillation Index and u
Summary
A large fraction of southern hemisphere glaciers is located in the Andes of South America (Pfeffer and others, 2014). Since local geographic conditions are implicitly included in empirical models, it removes the need to explicitly parameterise those for the location of the investigated glaciers This approach requires sufficiently long and temporally overlapping measurements of atmospheric predictors and glacier mass-balance in order to adequately capture robust empirical transfer functions between them. It requires knowledge of glacier-climate interactions in the region to ensure that physically meaningful predictors are chosen and captured empirical relationships represent underlying physical processes. ESD-based empirical glacier massbalance models (EGMs) are trained for individual glaciers They capture the transfer functions between glacier mass-balance variations and large-scale climate variability obtained from ERA-Interim reanalysis. To generate, evaluate and apply EGMs for each Andean glacier that is suitable for an ESD-based modelling approach
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
