Peter Kuma

Science and Software

The near-surface environment of the Southern Ocean is subject to particular biases in weather and climate simulations, particularly during the summer season, and relatively few analyses of cloud and radiation properties have been reported for the region. Here we provide an analysis of ship-based measurements of downwelling radiation, cloud fraction and cloud base height from the RSV Aurora Australis during the Kerguelen Axis marine science campaign which was conducted in the Southern Ocean south-east of the Kerguelen Plateau between January and March 2016. Our study period focussed on a 22-day interval during the first two months of the campaign. We compared estimates of cloud fraction obtained with a cloud imager and ceilometer, and found good agreement between the two measurement types, particularly when the camera images were analysed in a narrow overhead field to account for differences in the measurement techniques. We used the Interim European Centre for Medium-Range Weather Forecasts Reanalysis (ERA-Interim) and the Antarctic Mesoscale Prediction System Polar Weather and Research Forecasting model (Polar WRF) to provide comparison data for our measurements. We found that both comparison data sets generally underestimated cloud cover (observed cloud fraction ~0.96 compared with 0.87 for ERA-Interim and 0.63 for Polar WRF). As a consequence, the comparison data showed biases in both the surface shortwave irradiance (+59 W m⁻² for ERA-Interim and +154 W m⁻² for Polar WRF) and the longwave irradiance (−23 W m⁻² for ERA-Interim and −46 W m⁻² for Polar WRF). The observed mean net surface cloud radiative effect (CRE) of −228 W m⁻² was significantly more negative than found in previous observations in the Southern Ocean region, and compares with a net surface CRE of − 138 W m⁻² for ERA-Interim which also showed relatively strong cloud forcing. The observed net surface CRE bias for ERA-Interim of + 90 W m⁻² appears primarily the result of the reanalysis underestimating the cloud fraction, which at least partly relates to a lack of low clouds. Polar WRF was also found to have a deficit of low clouds. We characterised the relationship between the ratio of irradiances by Photosynthetically Active Radiation (PAR) and shortwave radiation and cloud transmittance. As a consequence of cloud, light levels were estimated as being below the level for light-limited photosynthesis during 31% of the available time the sun was above the horizon (69% of each day on average), compared with the expected clear-sky value of 10%. Over the campaign period, the Indian Ocean sector of the Southern Ocean was influenced by the positive phase of the Southern Annular Mode (SAM). Notably, the surface SAM index in January and March was the most positive observed since 1957. This situation generally led to near-surface climatological differences over the southern part of the campaign region over much of the period, which included significant negative anomalies in mean sea level pressure and air temperature, and positive anomalies in zonal wind. Overall, the cloudiness of our study region appeared to be above average for the time of year, but we could not identify a clear cause for this in the prevailing climatic conditions. While the level of shortwave radiation was likely below average for the time of year, this deficit is not likely to have significantly impacted on photosynthesis in the mixed layer of the ocean.