Intercomparison of tropical-to-high latitude teleconnection effects on Antarctic ice sheet surface mass balance parameters on multiple timescales in several atmospheric re-analyses

Increasing attention is being given to effects of tropical variability on high latitude climate via atmospheric bridges. Observed relationships are not stable, and some inconsistencies exist in the literature. However, recent results from the Ice Sheet System Model (ISSM) suggest that understanding these relationships, and how they affect ice sheet surface mass balance (SMB), has relevance for error reduction in model simulations of ice sheet mass loss and may help to improve sea level projections. Unfortunately, the lack of data at high latitudes means researchers must rely on atmospheric reanalyses, which are not always consistent with one another. We present examples from an intercomparison of how several existing atmospheric reanalysis products represent variability in precipitation and other atmospheric parameters over the Antarctic ice sheet that is believed to be linked to ENSO, the Madden-Julian Oscillation (MJO) and other aspects of tropical variability. Preliminary results show statistical correlations between precipitation averaged over the West Antarctic Ice Sheet (WAIS) and ENSO are very similar among products, but that notable differences exist between products in terms of correlations between ENSO and precipitation over the Antarctic Peninsula and East Antarctica. Primary EOF modes for both interannual and intraseasonal precipitation variability show notable differences between products. Overall, caution must be taken when drawing conclusions about tropical-high latitude teleconnections from currently available atmospheric reanalyses. This work supports a broader effort under development at NASA's Jet Propulsion Laboratory geared toward improving understanding and simulation of polar ice sheet outfluxes due to SMB using a coupled model framework involving JPL's Ice Sheet System Model (ISSM) and the Goddard Earth Observing System Model 5. This work also supports the need for continuing/improving polar observations.