Premium
Comparison of ECMWF analysis and forecast humidity data with CARIBIC upper troposphere and lower stratosphere observations
Author(s) -
Dyroff Christoph,
Zahn Andreas,
Christner Emanuel,
Forbes Richard,
Tompkins Adrian M.,
van Velthoven Peter F. J.
Publication year - 2014
Publication title -
quarterly journal of the royal meteorological society
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.744
H-Index - 143
eISSN - 1477-870X
pISSN - 0035-9009
DOI - 10.1002/qj.2400
Subject(s) - stratosphere , troposphere , extratropical cyclone , tropopause , environmental science , atmospheric sciences , water vapor , relative humidity , climatology , humidity , northern hemisphere , middle latitudes , atmosphere (unit) , latitude , polar vortex , meteorology , geology , geography , geodesy
Water vapour, relative humidity and temperature in the Northern Hemisphere extratropical upper troposphere/lowermost stratosphere (UT/LMS) of the operational European Centre for Medium‐Range Weather Forecasts (ECMWF) analysis and forecast system are compared with in situ measurements from Civil Aircraft for the Regular Investigation of the atmosphere Based on an Instrument Container (CARIBIC) passenger aircraft flights between 2005 and 2012. Modelled and observed temperatures are found to be in good agreement, with a small tendency towards a cold model bias of 0.45 K at higher temperatures. Water vapour and relative humidity also agree well in the upper troposphere. However, in the dry extratropical lowermost stratosphere, both analysis and forecast show a considerable moist bias. About 2 km above the tropopause, the specific humidity is overestimated, typically by 100–150% in summer and autumn as well as 50–100% in winter and spring. Five‐day backward trajectories indicate that the moist bias is associated with airmasses that have resided approximately at the aircraft flight level of around 230 hPa in high northern latitudes. Fast ascending airmasses from low latitudes show considerably less bias. There are indications that part of the model bias is due to small‐scale stratospheric intrusions that are unresolved by the model but are a frequent observed feature in the UT/LMS. Numerical diffusion of water vapour across the hydropause from the advection scheme and a lack of a constraint on humidity in the stratosphere in the analysis are also possible reasons for the moist bias. About 6% of all air masses transected by the CARIBIC aircraft are supersaturated with respect to ice. Previous changes in the operational analysis and forecast systems have considerably improved the representation of ice supersaturation, yet there is an indication that both still underestimate the occurrence of ice supersaturation measured during CARIBIC flights.