Homogenous atmospheric observations, state and flux quantities of the past 75 years

The proposed project intends (i) to contribute to improved quality of global reanalyses by providing homogeneity adjustments for upper air data back to the 1930s and (ii) to draw on the improved quality when evaluating global budget quantities from reanalyses. Both aims are core goals of phase II of the World Climate Research Programme (WCRP) Global Energy andWater cycle Experiment (GEWEX). The results will also contribute to the goals of the Global Monitoring for Environment and Security (GMES) programme, a major initiative for building a European capacity for Earth Observation. The homogenized radiosonde temperature products developed in the two preceding FWF projects have become the standard radiosonde temperature bias correction input for all major ongoing reanalyses (the ERA-Interim reanalysis, NASA`sModern Era Retrospective Re-Analysis (MERRA) and the new Japanese reanalysis (JRA-55). Yet it needs to be extended substantially backwards to before 1958 and also towards new parameters, most notably humidity in order to keep its world leading position. Within a still running FWF project the technical framework for homogenization of these data has been set up but only the feasibility of the approach could be demonstrated. A mature implementation that works for most of the early radiosonde records is, however, needed for reanalysis efforts planned in the near future, most notably within the EC framework project ERA-CLIM. This will be one major task of the proposed project. The improved observations should lead to improved reanalyses and these should lead to more reliable atmospheric energy and moisture budgets as well. It is important to investigate the low frequency variability not only of essential climate variables such as temperature but also of energy and moisture fluxes. Characteristic energy flow patterns during strong excursions of the Southern Oscillation Index have been diagnosed recently in the satellite era (1979-). More homogeneous input data and improved data assimilation methods should yield more reliable diagnostics back as far as to the early 1940s, when a strong El Nino event was likely related to the extreme northeastern European winters during World War II. The uncertainty of these diagnosed fluxes and their variability needs to be carefully assessed and reduced to be useful for validation of climate models. We expect the resulting homogenized datasets to be used as in situ upper air input for future reanalyses and to be available via public climate data portals. The project would support the aims of ERA-CLIM (where the PI is project partner) and of a new EC framework project to be submitted in autumn 2012. The compatibility of the project aims with international research programs and close cooperation with leading research centres, most notably the European Centre for Medium-range Weather Forecasts (ECMWF) and the National Center for Atmospheric Research (NCAR) should ensure high scientific impact of the proposed work.

Earths climate and its change can be characterized by state quantities (such as temperature) and by flux quantities, e.g. top of atmosphere radiation flux, precipitation or poleward horizontal energy transports. State and flux quantities are related to each other via the energy and water budgets. Flux quantities are as important as state quantities but they are more difficult to measure or must be calculated from state quantities. Changes in the fluxes as well as state quantities can be diagnosed only with considerable care despite modern observation systems because there are serious homogeneity issues. A data set is called homogeneous, if several decades or centuries old observations means can be strictly compared with those made today using modern measurement equipment. This is only rarely the case since early observations tend to have strong biases that need to be corrected. Without correction these observations can lead to completely wrong estimates of climate change. In this now concluding project measurements by radiosondes and tracked balloons (mostly temperature, wind and humidity) have been checked for outliers and inhomogeneities. The detected inhomogeneities have been adjusted and the adjustments have been made available to the scientific community. The global homogenized radiosonde wind data set GRASP developed in this project is the first of its kind. The homogeneity adjustments will improve the next comprehensive assessments of the global climate state, so called re-analyses. Those will allow more accurate estimates of climate change. The project has not only contributed to the improvement of reanalyses. It also used reanalysis data to quantify changes in global and regional energy and water budgets. A strong energy exchange between the Atlantic and Pacific Ocean basins modulated by the El Niño-Southern Oscillation signal could be detected and quantified. Contrary to the traditional paradigm it is much stronger than the El-Niño induced radiation of excessive heat to space. The budgets are also sensitive indicators of inconsistencies between data sets and could be used to detect inhomogeneities in the reanalysis data. This is the first step towards removing them in the next generation of reanalyses.