Towards improving the representation of polar regions in global reanalyses at the NASA GMAO

Abstract:

Atmospheric reanalyses have been employed in a wide variety of studies investigating processes in the changing Arctic and Antarctic. However reanalyses underperform in polar regions due in part to limitations in the background model, which are associated with spatial resolution and with inadequate representations of physical processes. Some chronic issues include the use of incorrect topography over ice sheets, persistent temperature biases, and discontinuities in time series due to abrupt changes to the observing system. These limitations provoke caution in the use of global reanalyses in polar regions, but also call into question their general applicability to evolving polar climate topics. For example, the assessment of conditions over Greenland and Antarctica in the most recent IPCC report relied heavily on limited area models rather than global reanalyses.

The performance of contemporary reanalyses over the north and south polar caps are reviewed with emphasis on differences in atmospheric energy and moisture budgets. In particular, the recently released MERRA-2 reanalysis is critically assessed. Significant changes that are relevant to polar regions in MERRA-2 over the previous MERRA version include an improved representation of glaciated land surfaces and sea ice albedo. Ongoing work to address issues of increased spatial resolution and the introduction of more complex physical processes related to surface hydrology over polar ice sheets will be described.