In the present study, the quantile-quantile mapping transformation of Amengual et al. (2012) is used to bias correct modelled daily maximum and minimum temperatures. This method amends errors in the mean, variability and shape of the cumulative distribution functions (CDFs) of the variables of interest. The overall adjustment process is shown in equation (S1.1) and is illustrated in Figure S1 for daily maximum temperatures. In equations (S1.1) to (S1.4), the ith ranked values are shown: bc i is the bias-corrected model data for some future period, o i are the observations, and s bi and s fi are the (uncorrected) model data for the baseline and future time periods respectively. The time periods of the modelled baseline and observations are the same.

Abstract. Meteorological and atmospheric models are being extended up to 80 km altitude but there are very few observing techniques that can measure stratospheric-mesospheric winds at altitudes between 20 and 80 km to verify model datasets. Here we demonstrate the feasibility of horizontal wind profile measurements using ground-based passive millimetre-wave spectroradiometric observations of ozone lines centred at 231.28, 249.79, and 249.96 GHz. Vertical profiles of horizontal winds are retrieved from forward and inverse modelling simulations of the line-of-sight Dopplershifted atmospheric emission lines above Halley station (75 • 37 S, 26 • 14 W), Antarctica. For a radiometer with a system temperature of 1400 K and 30 kHz spectral resolution observing the ozone 231.28 GHz line we estimate that 12 h zonal and meridional wind profiles could be determined over the altitude range 25-74 km in winter, and 28-66 km in summer. Height-dependent measurement uncertainties are in the range 3-8 m s −1 and vertical resolution ∼ 8-16 km. Under optimum observing conditions at Halley a temporal resolution of 1.5 h for measuring either zonal or meridional winds is possible, reducing to 0.5 h for a radiometer with a 700 K system temperature. Combining observations of the 231.28 GHz ozone line and the 230.54 GHz carbon monoxide line gives additional altitude coverage at 85 ± 12 km. The effects of clear-sky seasonal mean winter/summer conditions, zenith angle of the received atmospheric emission, and spectrometer frequency resolution on the altitude coverage, measurement uncertainty, and height and time resolution of the retrieved wind profiles have been determined.

Abstract. Meteorological and atmospheric models are being extended up to 80 km altitude but there are very few observing techniques that can measure stratospheric-mesospheric winds at altitudes between 20 km and 80 km to verify model data-sets. Here we demonstrate the feasibility of horizontal wind profile measurements using ground-based passive millimetre-wave spectroradiometric observations of ozone lines centred at 231.28 GHz, 249.79 GHz, and 249.96 GHz. Vertical profiles of horizontal winds are retrieved from forward and inverse modelling simulations of the line-of-sight Doppler-shifted atmospheric emission lines above Halley station (75°37’S, 26°14’W), Antarctica. For a radiometer with a system temperature of 1400 K and 30 kHz spectral resolution observing the ozone 231.28 GHz line we estimate that 12hr zonal and meridional wind profiles could be determined over the altitude range 25–74 km in winter, and 28–66 km in summer. Height-dependent measurement uncertainties are in the range 3–8 m s–1 and vertical resolution ~8–16 km. Under optimum observing conditions at Halley a temporal resolution of 1.5hrs for measuring either zonal or meridional winds is possible, reducing to 0.5hr for a radiometer with a 700 K system temperature. Combining observations of the 231.28 GHz ozone line and the 230.54 GHz carbon monoxide line gives additional altitude coverage at 85±12 km. The effects of clear-sky seasonal mean winter/summer conditions, zenith angle of the received atmospheric emission, and spectrometer frequency resolution on the altitude coverage, measurement uncertainty, and height and time resolution of the retrieved wind profiles have been determined.