Atmospheric Vertical Structure and Trends in Climate Data

Observations of the Earths surface temperature provide undeniable evidence of a changing climate. While surface temperature trends are in accordance amongst different groups, there are still unresolved issues regarding upper-air climate trends. Though overall agreement on a global warming of the troposphere and a cooling of the stratosphere is given, the uncertainty in trend rates and their vertical structure is large and limits the ability to draw robust and consistent inferences about climate trends. This is stated as a key issue in the recent IPCC report implying the need for data with better accuracy. Addressing this need is challenging since uncertainties exist in observations, reanalyses, and climate model output. Observations from weather satellites and balloons have several shortcomings since they were not intended to serve climate monitoring needs, which demand accurate and long-term stable measurements. In atmospheric reanalyses observations are integrated into models, containing effects of both observation and model errors. In current climate models discrepancies to observations are evident in vertical thermal structure and trends, especially in the upper troposphere and stratosphere. Radio Occultation (RO) provides independent observations with beneficial characteristics in this context. The traceability to time measurements with precise atomic clocks assures a long-term stable and consistent data record with global coverage. Accuracy, low structural uncertainty, and use for climate studies have been demonstrated. High quality and vertical resolution offer the distinct advantage for assessing the vertical thermodynamic structure. The central aim of the project VERTICLIM is the exploration and evaluation of the vertical structure of atmospheric climate variability and climate trends, their regional imprints, and relevant processes from the surface to the stratopause. Focus periods are 20022015 with dense data set coverage for rigorous short-term study, and 19792015 with good coverage for complementary longer-term study. New insights will be gained on recent climatic changes in the troposphere and stratosphere by systematically exploiting upper-air records from observations and reanalyses, and climate models. Evaluating the RO record as reference and the inter-consistency of other high-quality observations such as MIPAS, SABER, radiosondes, and of AMSU/SSU bulk temperatures, will provide essential new information on the climate quality of upper-air observations. Exploration of atmospheric key characteristics, including annual cycle, tropopause/stratopause dynamics, and climate variability modes will reveal key strengths, weaknesses, and improvement potential of reanalyses and models. Also new insights into the vertical structure of trends and their regional imprints in the atmosphere compared to the surface will be gained from RO and best-evaluated records. Overall we aim with VERTICLIM to draw a picture of unprecedented quality and rigor of the vertical structure of atmospheric climate variability and trends and to reveal key skills of upper-air observations, reanalyses, and climate models with RO as reference climate data record.

The Earth's atmosphere is governed by a range of natural variability modes, which need to be well characterized to be able to accurately separate them from anthropogenic climate trends. Precise observations for climate monitoring are therefore required, meeting the quality criteria of the Global Climate Observing System (GCOS) program. In this respect, Radio Occultation (RO) measurements based on Global Navigation Satellite Systems (GNSS) signals have the beneficial properties of long-term stability, all-weather capability, global coverage, high accuracy and vertical resolution in the upper troposphere and lower stratosphere. Here we present results on the characterization of the vertical thermal structure of variability and trends, and relevant processes from the surface to the stratosphere, using RO and other upper-air observations in comparison with reanalyses and climate models. We confirm the climate quality of RO observations by quantifying the structural uncertainty of multi-satellite RO records from different processing centers. RO records are consistent within 8 km to 25/30 km at low to mid-latitudes. Structural uncertainty for temperature is <0.05 K per decade nearly everywhere in this region, and <0.1 K per decade including high latitudes. The maturity of RO has further improved towards a climate data record meeting the GCOS criteria. We made the Wegener Center RO OPSv5.6 record publicly available for use in climate and other high-accuracy applications. We provide novel altitude-resolved atmospheric variability proxies constructed directly from RO temperature measurements. They represent main atmospheric modes such as El Niño-Southern Oscillation (ENSO) and Quasi-Biennial Oscillation (QBO), and are greatly beneficial in trend regression analyses. Furthermore, we investigate various other sources of variability such as tropopauses, gravity waves, or volcanic explosive eruptions. We find substantial improvement in the agreement of temperature trends from reprocessed stratospheric observations with chemistry-climate model data, showing that the stratospheric cooling over the satellite era weakens after 1998. Vertically resolved trends from RO in the period 2001 to 2018 show warming in the whole troposphere and above the tropical tropopause, while cooling is observed in the tropical stratosphere. Arctic amplification is seen to affect the equator-to-pole temperature differences. Trends from RO are found largely consistent with those from radiosondes and reanalyses, though some differences are revealed. The results have been provided for the upcoming Sixth Assessment Report (AR6) of the Intergovernmental Panel on Climate Change (IPCC). RO as a unique data source enables an improved representation of the vertical thermal structure and atmospheric trends. This is crucial for studies of atmospheric physics and dynamics, climate trend detection, and the evaluation of next-generation climate models.