Estimating radar reflectivity - Snowfall rate relationships and their uncertainties over Antarctica by combining disdrometer and radar observations

N Souverijns,A Gossart,S Lhermitte,I.V Gorodetskaya,S Kneifel,M Maahn,F.L Bliven,N.P.M Van Lipzig

doi:10.1016/j.atmosres.2017.06.001

Abstract

Snowfall rate (SR) estimates over Antarctica are sparse and characterised by large uncertainties. Yet, observations by precipitation radar offer the potential to get better insight in Antarctic SR. Relations between radar reflectivity (Ze) and snowfall rate (Ze-SR relations) are however not available over Antarctica. Here, we analyse observations from the first Micro Rain Radar (MRR) in Antarctica together with an optical disdrometer (Precipitation Imaging Package; PIP), deployed at the Princess Elisabeth station. The relation Ze=A*SRB was derived using PIP observations and its uncertainty was quantified using a bootstrapping approach, randomly sampling within the range of uncertainty. This uncertainty was used to assess the uncertainty in snowfall rates derived by the MRR. We find a value of A = 18 [11–43] and B = 1.10 [0.97–1.17]. The uncertainty on snowfall rates of the MRR based on the Ze-SR relation are limited to 40%, due to the propagation of uncertainty in both Ze as well as SR, resulting in some compensation. The prefactor (A) of the Ze-SR relation is sensitive to the median diameter of the snow particles. Larger particles, typically found closer to the coast, lead to an increase of the value of the prefactor (A=44). Smaller particles, typical of more inland locations, obtain lower values for the prefactor (A=7). The exponent (B) of the Ze-SR relation is insensitive to the median diameter of the snow particles. In contrast with previous studies for various locations, shape uncertainty is not the main source of uncertainty of the Ze-SR relation. Parameter uncertainty is found to be the most dominant term, mainly driven by the uncertainty in mass-size relation of different snow particles. Uncertainties on the snow particle size distribution are negligible in this study as they are directly measured. Future research aiming at reducing the uncertainty of Ze-SR relations should therefore focus on obtaining reliable estimates of the mass-size relations of snow particles.

Highlights

The Antarctic Ice Sheet (AIS) is the largest ice body on earth, having a volume equivalent to 58.3 m global mean sea level rise (Vaughan et al, 2013)
As the Precipitation Imaging Package (PIP) measures the particle size distribution (PSD) directly, this uncertainty term is limited to measurement errors of the instrument
Using the Precipitation Imager Package (PIP) and a Micro Rain Radar (MRR), a Ze-snowfall rate (SR) relation (Ze = A*SRB) over Antarctica was derived by performing bootstrapping simulations taking different uncertainty terms into 655 account

Summary

Introduction

The Antarctic Ice Sheet (AIS) is the largest ice body on earth, having a volume equivalent to 58.3 m global mean sea level rise (Vaughan et al, 2013). Precipitation is the dominant source term in the surface mass balance of the AIS This quantity is not well constrained in both models and observations (Bromwich et al, 2004; Palerme et al, 2014). Direct observations over the AIS are not coherent, as they are sparse in space and time and since acquisition techniques differ. These records are usually determined from ice cores, satellite products or stake measurements. Observations are often 15 disturbed by blowing snow, which makes the distinction between transported and precipitating snow impossible (Knuth et al, 2010) This impedes the use of precipitation gauges over Antarctica, as blowing snow may enter the gauge, while high wind speeds may lead to an undercatchment of precipitation (Yang et al, 1999). Precipitation observations stay mostly limited 20 to continent-wide averages (e.g. Vaughan et al, 1999)

Objectives

Methods

Results

Conclusion