Title: North Atlantic Oscillation response to Labrador Sea Ice
Start Date: January 10, 2000
Summary:This experiment has been performed in oprder to explore the atmospheric response to sea ice anomalies in the Labrador Sea.
The model employed in this study is the ARPEGE/IFS model, used by Météo-France and documented in Déqué (1994), Doblas-Reyes et al. (1998). The horizontal resolution used here is a linear T63 truncation (T63L), which has a lat/lon grid spacing of about 2.8° everywhere on the globe. There are 31 levels in the vertical, with 20 in the troposphere and 10 in the stratosphere.
The sea-ice boundary is defined in ARPEGE by the -1.9°C SST isotherm, which means that sea-ice can be perturbed by perturbing the SSTs. A localized SST anomaly, centred in the Labrador Sea, was constructed by taking a 16 year long monthly mean SST dataset for the period 1979 to 1995 from the AMIP project Gates (1992). On basis of this dataset, we have formed an index of area averaged SSTs for the calendar months November through March. The annual cycle was removed by subtracting the long term mean of each calendar month and then the SSTs were averaged over the area 60°W - 50°W, 55°N - 60°N in the Labrador Sea. This procedure resulted in a Labrador Sea SST time series index (not shown). An SST anomaly was subsequently obtained by regressing global SST anomalies onto the index. The amplitude of the anomaly decreases with distance out from its centre (index area). To prevent weak remote anomalies, the anomaly has been set to zero when the correlation coefficient (between the index and the global anomalies) is less than r = 0.34. The remaining anomaly was multiplied with ±3, corresponding to ±3 standard deviations of the SST index and superposed on the January, February and March climatological SST dataset (Reynolds and Smith 1994) (Figure 1). The SST index spans 5 standard deviations during these 16 analysed winters (80 realizatons).
The model experiments were designed as follows. First, one 14 year control simulation (CTRL) was run with climatological, seasonally varying SSTs. The annual cycle of the SSTs is constituted by monthly means that are interpolated linearly in time and updated daily. From CTRL, 14 initial states from 14 different November months were extracted. For each initial state, two winter runs (Nov-Mar) were made, one with maximal (LABMAX) and one with minimal (LABMIN) Labrador Sea ice-cover.
MORE EXPERIMENTS
|
