The vertical thermal structure of the atmosphere reflects a balance between radiative, convective, and dynamical heating and cooling processes of the surface-atmosphere system. In the troposphere radiative processes involving greenhouse gases, aerosols, and clouds dominate together with strong moist convection and dynamical (vertical) motions (e.g., Holton 2004). The latter are weak in the stratosphere where radiative heating and cooling rates due to mainly carbon dioxide, ozone, and water vapor are of importance (e.g., Andrews et al. 1987). Changes in the upper troposphere-lower stratosphere (UTLS) region have strong impact on the Earth's climate system, consequently climate monitoring is in need of long-term stable, self-calibrating, and well height-resolved global data in this region. Up to now, temperature changes are commonly used as climate change benchmarks in trend studies, mostly with

Wegener Center for Climate and Global Change (WegCenter) and Institute for Geophysics, Astrophysics, and Meteorology (IGAM), University of Graz, Austria e-mail:[email protected]

focus on surface changes. The establishment of radiosondes in the 1960 s and space borne measurement systems in the late 1970 s enabled the investigation of upper-air temperature (Karl et al. 2006).

Climatologies based on atmospheric profiles retrieved from RO observations allow the monitoring and diagnosis of climate change (Leroy et al. 2006). The RO method provides high quality atmospheric parameters in the UTLS comprising refractivity, pressure, geopotential height, temperature, and in the lower to middle troposphere specific humidity (see e.g. for CHAMP data Wickert et al. 2004; Hajj et al. 2004; Steiner et al. 2006). These parameters are highly relevant to investigate upper tropospheric warming and lower stratospheric cooling in a changing climate.

This study aims at demonstrating the potential of the whole set of RO accessible parameters as climate change indicators in the UTLS region. We define an indicator as a variable, which succeeds best to map the process of anthropogenic climate change in a certain space and time domain. Due to the still limited length of real RO time series (end 2001-2007), we use climate simulations of three representative global circulation models (GCMs). Section 2 gives an overview of the data used within this study and their characteristics. The study design and the trend analysis method as well as the method used to identify the most sensitive parameters and their respective temporal and spatial features is described in Sect. 3. Results are presented in Sect. 4 and discussed with respect to RO in Sect. 5, while in Sect. 6 a summary is given and conclusions are drawn.

0 0

Post a comment