An Assessment of Numerical Weather Prediction–Derived Low-Cloud-Base Height Forecasts

Abstract

Abstract As demand for flight operations in Antarctica grows, accurate weather forecasting of cloud properties such as extent, cloud base, and cloud-top altitude becomes essential. The primary aims of this work are to ascertain relationships between numerical weather prediction (NWP) model output variables and surface-observed cloud properties and to develop low-cloud-base (<2000 m) height prediction algorithms for use across Antarctica to assist in low-cloud forecasting for aircraft operations. NWP output and radiosonde data are assessed against surface observations, and the relationship between the relative humidity RH profile and the height of the observed low-cloud base is investigated. The ability of NWP-derived RH and ice–water cloud optical depth profiles to represent the observed low-cloud conditions around each of the three Australian stations in East Antarctica is assessed. NWP-derived RH is drier than that reported by radiosonde from ground level up to ~2000 m. This trend reverses in the higher troposphere, and the largest positive difference is observed at ~10 000 m. A consequence is very low RH thresholds are needed for low-cloud-base height prediction using NWP RH profiles. RH and optical depth–based threshold techniques all show skill in reproducing the observed cloud-base height at all Australian Antarctic stations, but the radiosonde-derived RH technique is superior in all cases. This comparison of three low-cloud-base height retrieval techniques provides the first documented assessment of the relative efficacy of each technique in Antarctica.