Impact of the additional FASTEX radiosonde observations on the high‐resolution limited‐area model (HIRLAM) data‐assimilation and forecasting system

Abstract

AbstractDuring the Fronts and Atlantic Storm‐Track EXperiment (FASTEX) the radiosonde network was considerably enhanced over the North Atlantic area, making this an ideal period for assessing the impact of additional radiosonde observations, in data‐sparse regions, on the atmospheric data‐assimilation and numerical weather‐forecast systems.We have carried out an observing‐system experiment (OSE) with a limited‐area data‐assimilation system. Two parallel experiments have been run, one with and one without the extra radiosonde data.We have used both observation verification and field (or analysis) verification to compare the analyses and forecasts from the parallel experiments. On a daily basis, the extra data can have both positive and negative impacts on the different meteorological variables considered here. However, averaged over the whole period the impact is definitely positive, especially on upper model levels. Our results strongly support the inclusion of extra radiosonde data, like those provided by FASTEX. They could be of significant benefit in reducing the errors in operational numerical weather‐prediction products if they were available routinely.