The vertical structure of the lower Arctic troposphere analysed from observations and the ERA‐40 reanalysis

Abstract In situ atmospheric observations in the central Arctic are few and mostly from near the surface. A majority are from coastal regions whereas soundings over the Arctic Ocean are rare. This limits our understanding of the Arctic atmosphere, in particular aloft. It has been established that the vertical thermal structure is often stably stratified; this has been termed the ‘Arctic inversion’. It has also been established that near‐surface warming in the Arctic has been larger than the global mean warming during the last several decades. To estimate climate trends in this data‐sparse region, reanalysis data have often been used. In this paper we analyse the vertical thermal structure of the lower troposphere over the Arctic Ocean, using soundings from the SHEBA project. We find a strong annual cycle with strong surface inversions occurring only during autumn and winter, typically 500–800 metres deep and ∼10 °C strong. Summer is dominated by weaker elevated inversions at ∼100–400 m, a few hundred metres deep. Interestingly, this latter type of inversion also occurs frequently in winter, almost half the time. These soundings thus indicate that associating Arctic winter only with strong surface inversions is not entirely correct. We also compare these soundings to the ERA‐40 reanalysis data. Systematic biases in ERA‐40 in the SHEBA region include a near‐surface warm bias, on average ∼0.5–1.0 °C, and a slight mid‐troposphere cool bias. There is a significant difference in ERA‐40 performance statistics for the SHEBA year comparing with years without soundings for the same region. The analysis increment—a measure of the impact of the observations in the assimilation process—confirms this. For example, the assimilation of the SHEBA soundings reduces the near‐surface warm bias by about 50%. However, the overall vertical structure and its annual variation are surprisingly insensitive to the assimilation of the soundings, and are in fact well represented by ERA‐40. We speculate that the main improvement in assimilating the SHEBA soundings lies in an improvement in the timing of weather systems whereas their climatological vertical structure is less affected. Copyright © 2009 Royal Meteorological Society