The accuracy of stratospheric analyses in the northern hemisphere inferred from long‐duration balloon flights

Abstract

AbstractIn January and February 2002, six superpressure balloons (SPBs) were launched from Kiruna, Sweden (69°N,21°E). The balloons drifted in the polar lower stratosphere for up to 45 days. Temperature, wind and altitude observations were collected every 15 minutes during the flights. This dataset serves as a reference to assess the accuracy of the European Centre for Medium‐Range Weather Forecasts (ECMWF) analyses and the National Centers for Environmental Prediction/National Center for Atmospheric Research (NCEP/NCAR) reanalyses in the lower stratosphere during the 2002 winter. In particular, it is found that both models succeeded in simulating the large temperature fluctuations observed during the SPB flights, which were induced by a geographical shift between the stratospheric vortex and the low temperature area. Furthermore, the model biases are found to be small for all variables (temperature, wind, height). Nevertheless, it is found that NCEP/NCAR reanalyses tend to be slightly warmer (0.8 K) than the observations, while the converse is true for ECMWF analyses (−0.3 K). The observations also exhibit small‐scale fluctuations that are presumably produced by (inertia‐)gravity waves and which induce some scatter in the analysis/observation comparisons. Finally, trajectory comparisons are performed. It is found that trajectories built with ECMWF winds are more accurate than those built with NCEP/NCAR winds. With both models, the trajectory errors increase with time upto about 15 days and then tend to fluctuate. Typical errors after 15 days are 1000 ± 1200 km for ECMWF and 2300 ± 1300 km for NCEP trajectories. Copyright © 2004 Royal Meteorological Society