<p>The global hydrological cycle is expected to intensify under a warming climate. Since extratropical Rossby wave trains are triggered by tropical convection, this will impact the atmospheric circulation in the extratropics. Owing to the approximate linearity of the teleconnection pattern, we can use a method based in linear response theory to quantify this extratropical response using a step response function. We have examined the step response functions for a selection of CMIP5 pre-industrial control runs and reanalysis data, &#160;in particular studying the response during the boreal winter. We found there to a large intermodel spread in the response pattern owing to differences in representations of the model basic state. In the current work, we use a 'perfect model' approach to conduct a systematic study of the performance of the linear response method in projecting future winter-time northern hemisphere circulation changes using the present day (1986-2005) model basic states, comparing these to those projected by CMIP5 models under a 3 degree rise in mean global temperature anomaly above pre-industrial. We demonstrate how, given a projected precipitation change pattern, the linear response theory method can compete with the models in providing faithful projections for the extratropical circulation changes.</p>
Teoria reakcji na pozycję w podejściu modelowym w taksonomii / Item response theory in model-based clustering
