Can a climate model reproduce extreme regional precipitation events over England and Wales?

The ability of the HiGEM climate model to represent high-impact, regional, precipitation events is investigated in two ways. The first focusses on a case study of extreme regional accumulation of precipitation during the passage of a summer extra-tropical cyclone across southern England on 20 July 2007 that resulted in a national flooding emergency. The climate model is compared with a global Numerical Weather Prediction (NWP) model and higher resolution, nested limited area models. While the climate model does not simulate the timing and location of the cyclone and associated precipitation as accurately as the NWP simulations, the total accumulated precipitation in all models is similar to the rain gauge estimate across England and Wales. The regional accumulation over the event is insensitive to horizontal resolution for grid spacings ranging from 90km to 4km. Secondly, the free-running climate model reproduces the statistical distribution of daily precipitation accumulations observed in the England-Wales precipitation record. The model distribution diverges increasingly from the record for longer accumulation periods with a consistent under-representation of more intense multi-day accumulations. This may indicate a lack of low-frequency variability associated with weather regime persistence. Despite this, the overall seasonal and annual precipitation totals from the model are still comparable to those from ERA-Interim.

The potential impact of changes in lower stratospheric water vapour on stratospheric temperatures over the past 30 years

This study investigates the potential contribution of observed changes in lower stratospheric water vapour to stratospheric temperature variations over the past three decades using a comprehensive global climate model (GCM). Three case studies are considered. In the first, the net increase in stratospheric water vapour (SWV) from 1980–2010 (derived from the Boulder frost-point hygrometer record using the gross assumption that this is globally representative) is estimated to have cooled the lower stratosphere by up to ∼0.2 K decade−1 in the global and annual mean; this is ∼40% of the observed cooling trend over this period. In the Arctic winter stratosphere there is a dynamical response to the increase in SWV, with enhanced polar cooling of 0.6 K decade−1 at 50 hPa and warming of 0.5 K decade−1 at 1 hPa. In the second case study, the observed decrease in tropical lower stratospheric water vapour after the year 2000 (imposed in the GCM as a simplified representation of the observed changes derived from satellite data) is estimated to have caused a relative increase in tropical lower stratospheric temperatures by ∼0.3 K at 50 hPa. In the third case study, the wintertime dehydration in the Antarctic stratospheric polar vortex (again using a simplified representation of the changes seen in a satellite dataset) is estimated to cause a relative warming of the Southern Hemisphere polar stratosphere by up to 1 K at 100 hPa from July–October. This is accompanied by a weakening of the westerly winds on the poleward flank of the stratospheric jet by up to 1.5 m s−1 in the GCM. The results show that, if the measurements are representative of global variations, SWV should be considered as important a driver of transient and long-term variations in lower stratospheric temperature over the past 30 years as increases in long-lived greenhouse gases and stratospheric ozone depletion.

Low rank representation on Riemannian manifold of symmetric positive definite matrices

Sparse coding aims to find a more compact representation based on a set of dictionary atoms. A well-known technique looking at 2D sparsity is the low rank representation (LRR). However, in many computer vision applications, data often originate from a manifold, which is equipped with some Riemannian geometry. In this case, the existing LRR becomes inappropriate for modeling and incorporating the intrinsic geometry of the manifold that is potentially important and critical to applications. In this paper, we generalize the LRR over the Euclidean space to the LRR model over a specific Rimannian manifold—the manifold of symmetric positive matrices (SPD). Experiments on several computer vision datasets showcase its noise robustness and superior performance on classification and segmentation compared with state-of-the-art approaches.