Since the upper troposphere-lower stratosphere (UTLS) region reacts sensitively to climate change, the variations of its fundamental parameters such as temperature, geopotential height of pressure levels, and refractivity are promising candidates for monitoring the climate. All these key parameters are provided by RO observations with high quality in the UTLS and have high potential for climate analysis and monitoring (e.g., Leroy et al. 2006; Steiner et al. 2007; Foelsche et al. 2008a,b, 2009) and shall be explored to find the most favorable indicators for
Wegener Center for Climate and Global Change (WegCenter) and Institute for Geophysics, Astrophysics, and Meteorology (IGAM), University of Graz, Austria e-mail: [email protected]
monitoring global atmospheric change. Due to the limited time period of available RO data (available on a continuous basis only since September 2001) the parameters are explored in both reanalysis and global climate model (GCM) data first. Climate models such as the ECHAM5 of the Max-Planck-Institute for Meteorology (MPI-M) Hamburg provide long-term climate scenarios and are used over the time frame 1961-2064. Reanalysis data, e.g., the ERA-40 data set of the European Centre for Medium-Range Weather Forecasts (ECMWF), are used from 1961 onwards, with special focus on the time period after 1979 when satellite data were assimilated.
Complementary to classical trend testing methods to find the indicators of choice (Lackner et al. 2009), a novel approach to visually explore the climate data sets is used in this study. The interactive visual analysis tool SimVis (Doleisch et al. 2003; Kehrer 2007) has been developed with special focus on dealing with large data sets, which makes it particularly well applicable to the data fields occurring in climate research. In SimVis, different aspects of the whole data set can be concurrently analyzed in multiple linked views. The sophisticated feature specification tools of SimVis provide a way to interactively select regions of interest (the so-called features) in time and space. These techniques are used to gain an overview over all data sets, to easily reveal deficiencies in the data, and to localize regions of trends with high signal-to-noise ratio. No prior knowledge of the fields needs to be presumed, and no subset needs to be preselected, since the data can be explored as a whole at once. Interesting features can be specified while interactively exploring the field. These characteristics can be regarded as the main advantages compared to classical statistical methods.
Section 2 gives a brief description of the data sets used in this study. In Sect. 3 the SimVis software tool is presented and its application to climate data is explained. The results of the data set exploration are shown in Sect. 4, and conclusions are drawn in Sect. 5.
Was this article helpful?