The impact of varying atmospheric forcing on the thickness of Arctic multi-year sea ice
Date
2002
Authors
Dumas, Jacqueline Anne
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
The response of changing atmospheric forcing on multi-year sea ice in the Arctic is studied using a one-dimensional thermodynamic sea ice model forced with meteorological observations made at manned drifting stations during the 1954-91 period. These observations are first analyzed to explore variability and trends. We find that spring and fall wind speeds have increased by roughly +5 (+/-4) to +8 (+/-7) mis/century in most regions since 1974.
Model simulations indicate that, over the entire period, the only significant decreasing trend occurs in summer ice thickness in the Siberian portion of the Arctic Ocean (-0.06 +/- 0.06). From 1954-70, there are increasing trends in ice thickness in most seasons over the portion of the Arctic Ocean covered by drifting stations (+0. 13 +/- 0.1 m/decade), and in the Siberian Arctic (+0.25 +/- 0.1 m/decade). A dip in ice thickness takes place from 1971-77, in three of the four regions studied (Total, Central and North American Arctic), which corresponds to a period of positive AO phase, negative PNA phase and a shift in the ENSO cycle in 1976-77. Following this, from 1978-91 there are decreasing trends in the Total (0.30 +/- 0.2 m/decade) and Central Arctic (0.35 +/- 0.2 m/decade) in all seasons. This indicates that an important part of the variability and trends in ice thickness are thermodynamically driven. 10 to 20% of the variance in ice thickness is explained by the AO and PNA patterns. A positive phase in the AO leads to thinner ice due to surface air temperature increase, and a positive phase in the PNA leads to thicker ice due to stronger winds which increase the latent heat flux over the ice. These patterns have an impact on multi-year ice thickness, although not as dominant as in the atmosphere.
In sensitivity studies in which the albedo parameterization is altered, we see the signature of multiple equilibria in ice thickness, which indicates that within the uncertainty in the albedo parameterization the sea ice regime is close to the threshold of switching between equilibria. Therefore caution must be used in selecting the albedo parameterization to simulate the seasonal cycle of multi-year ice. Moreover if this model is correct, we may be in for an unexpected shock as climate warms.