Arctic boundar-layer inversions from SHEBA observations and ERA-40 reanalysis data

Previous analyses of historical radiosounding data from Russian ice-drift stations indicated the presence of quasi-persistent inversion conditions over the Arctic Ocean; this has lead to the concept of the “Arctic inversion”. Since the termination of these long-term observations, soundings in the central Arctic are scare. Meanwhile, the Arctic has been warming at a rate about twice the globally averaged warming. Several hypotheses have been postulated to explain this strong feedback. Common for most of these is that they relate to processes acting near the surface; the most well known is the so-called “ice-and-snow-albedo feedback”. As a result of this debate it becomes important to examine the vertical structure of the Arctic inversion. But this is hindered by the lack of real in-situ data, and reanalysis products such as the ERA-40 reanalysis from the European Center for Medium-Range Weather Forecasts (ECMWF) have become popular. With this also comes a need to evaluate such data in the Arctic.

This study first analyses the vertical structure of the lowest troposphere from soundings made during the SHEBA experiment. Temperature profiles from ERA-40 are then compared with the observations, both directly and in a “climatological” sense. One problem with such an evaluation is that the SHEBA soundings were also assimilated in the reanalysis; thus an agreement may be caused by this rather than by the quality of the reanalysis model. We employ several methods to analyze this effect: i) Comparing ERA-40 “first-.guess” profiles to the observations; ii) Comparing ERA-40 analysis to “first-guess” (the so-called analysis increment); iii) Comparing the annual climatology in ERA-40 for years without and with assimulations of soundings in the SHEBA region.

The results of this study are: 1) The lower Arctic troposphere is stably stratified in both winter and summer, although less so in the summer; 2) A significant annual cycle exist in inversion properties. Winter is characterized by strong inversions alternating between surface inversions and elevated inversions, while summer is characterized by weaker elevated inversions; 3) ERA-40 has a significant low-level warm bias and a smaller mid-troposphere cool bias and the assimilation of the SHEBA soundings has a significant impact on details in the reanalysis; 5) ERA-40 provides a surprisingly good description of the climatology of inversion characteristics.