Antarctica’s x-factor: How does Antarctic precipitation change with temperature?

Snowfall is by far the most important positive contributor to the overall mass balance of the Antarctic Ice Sheet, potentially buffering temperature-induced dynamical ice loss in a warming climate. Previous studies have proposed that Antarctic snowfall will increase along the Clausius-Clapeyron relationship, describing the saturation water vapour pressure as a function of temperature (7% change for 1°C of warming). Due to cold temperatures and continentality in the interior, this general, first-order explanation may not hold true for snowfall changes across the ice sheet. In this study, we investigate how this first-order approximation can be modified to more reliably represent snowfall changes in a warming climate for simulations of the Antarctic Ice Sheet.

To characterise the present-day precipitation pattern, we use reanalysis data and make use of state-of-the-art model data from the CMIP6 modelling project as well as regional model data. We analyse how the sensitivity of Antarctic precipitation to temperature changes is represented in models and how it potentially changes in the future. We use least-squares linear regression to determine the sensitivity factor, Antarctica’s x-factor, that is used in ice-sheet models to scale precipitation. 

With our statistical analyses, we show that sensitivities of column-integrated water vapour, precipitation, snowfall, net precipitation, and surface mass balance to temperature changes are fairly similar under present-day conditions; implying that the exponential relationship of saturation water vapour pressure to temperature could generally lead to additional mass gains of the Antarctic Ice Sheet with warming. However, we find that the relationship of Antarctic precipitation to temperatures across the ice sheet is not constant, but decreases with ongoing warming. Taking these changes into account could give a more reliable estimate of future precipitation changes than existing approaches. We demonstrate that a linear approximation of the exponential relationship between Antarctic precipitation and temperature becomes more and more imprecise in a warming climate, both for computing the sensitivity factor and to scale Antarctic precipitation in models.

We propose a new way to extract the sensitivity factor of Antarctic precipitation to temperature which takes regional variations and the temperature dependence into account. The temperature dependence becomes more important the higher the warming becomes. Considering local warming rates, we show the necessity of introducing a temperature-dependent scaling factor in ice-sheet models, especially for high-end or long-term sea-level projections.