Radar altimeters are important tools to monitor the volume of the ice sheets. The penetration of radar waves in the snowpack is a major source of uncertainty to retrieve surface elevation. To correct this effect, a better understanding of the sensitivity of the radar waveforms to snow properties is needed. Here, we present an extension of the Snow Model Radiative Transfer (SMRT) to compute radar waveforms and conduct a series of simulations on the Antarctic ice sheet. SMRT is driven by snow and surface roughness properties measured over a large latitudinal range during two field campaigns on the Antarctic Plateau. These measurements show that the snowpack is rougher, denser, less stratified, warmer, and has smaller snow grains near the coast than on the central Plateau. These simulations are compared to satellite observations in the Ka, Ku, and S bands. SMRT reproduces the observed waveforms well. For all sites and all sensors, the main contribution comes from the surface echo. The echo from snow grains (volume scattering) represents up to 40% of the amplitude of the total waveform power in the Ka band, and less at the lower frequencies. The highest amplitude is observed on the central Plateau due to the combination of higher reflection from the surface, higher scattering by snow grains in the Ka and Ku bands, and higher inter-layer reflections in the S band. In the Ka band, the wave penetrates in the snowpack less deeply on the central Plateau than near the coast because of the strong scattering caused by the larger snow grains. The opposite is observed in the S band, the wave penetrates deeper on the central Plateau because of the lower absorption due to the lower snow temperatures. The elevation bias caused by wave penetration into the snowpack show a constant bias of 10 cm for all sites in the Ka band, and a bias of 11 cm, and 21 cm in the Ku band for sites close to the coast and the central Plateau, respectively. Now that SMRT is performing waveform simulations, further work will address how the snowpack properties affect the parameters retrieved by more advanced retracking algorithms such as ICE-2 for different snow cover surfaces.

雷达高度计是监测冰盖体积的重要工具。雷达波在积雪中的穿透是获取地表高程的主要不确定性来源。为了纠正这种影响，需要更好地了解雷达波形对雪特性的敏感性。在这里，我们展示了雪模型辐射传输 (SMRT) 的扩展，以计算雷达波形并在南极冰盖上进行一系列模拟。SMRT 是由在南极高原的两次野外活动期间在大纬度范围内测量的雪和表面粗糙度特性驱动的。这些测量结果表明，与中部高原相比，海岸附近的积雪更粗糙、更密集、分层更少、更温暖，并且雪粒更小。这些模拟与 Ka、Ku、和 S 波段。SMRT 很好地再现了观察到的波形。对于所有站点和所有传感器，主要贡献来自表面回波。来自雪粒的回波（体积散射）代表了 Ka 波段中高达总波形功率幅度的 40%，在较低频率处更少。由于来自表面的更高反射、Ka 和 Ku 波段雪粒的更高散射以及 S 波段更高的层间反射的组合，在中央高原观察到最高振幅。在 Ka 波段，由于较大的雪粒引起的强烈散射，波在中央高原的积雪中的穿透深度小于海岸附近。在 S 波段观察到相反的情况，由于雪温较低，吸收率较低，波在中央高原穿透得更深。波浪穿透积雪引起的高程偏差显示，Ka 波段所有站点的恒定偏差为 10 厘米，而靠近海岸和高原中部的站点的 Ku 波段偏差为 11 厘米和 21 厘米，分别。现在 SMRT 正在执行波形模拟，进一步的工作将解决积雪特性如何影响由更高级的重新跟踪算法（如 ICE-2）针对不同积雪表面检索到的参数。