24 Jul 2023
 | 24 Jul 2023
Status: this preprint is open for discussion.

Machine learning of Antarctic firn density by combining radiometer and scatterometer remote sensing data

Weiran Li, Sanne B. M. Veldhuijsen, and Stef Lhermitte

Abstract. Firn density plays a crucial role in assessing the surface mass balance of the Antarctic ice sheet. However, our understanding of the spatial and temporal variations in firn density is limited due to i) spatial and temporal limitations of in situ measurements, ii) potential modelling uncertainties, and iii) lack of firn density products driven by satellite remote sensing data. To address this gap, this paper explores the potential of satellite microwave radiometer (SMISS) and scatterometer (ASCAT) observations for assessing spatial and temporal dynamics of dry firn density over the Antarctic ice sheet. Our analysis demonstrates a clear relation between density anomalies at a depth of 4 cm and fluctuations in satellite observations. However, a linear relationship with individual satellite observations is insufficient to explain the spatial and temporal variation of snow density. Hence, we investigate the potential of a non-linear Random Forest (RF) machine learning approach trained on radiometer and scatterometer data to derive the spatial and temporal variations in dry firn density. In the estimation process, ten years of SSMIS observations (brightness temperature), ASCAT observations (backscatter intensity), and polarisation and frequency ratios derived from SSMIS observations are used as input features to a random forest (RF) regressor. The regressor is first trained on time series of modelled density and satellite observations at randomly sampled pixels, and then applied to estimate densities in dry firn areas across Antarctica. The RF results reveal a strong agreement between the spatial patterns estimated by the RF regressor and the modelled densities. The estimated densities exhibit an error of ± 10 kg m−3 in the interior of the ice sheet and ± 20 kg m−3 towards the ocean. However, the temporal patterns show some discrepancies, as the RF regressor tends to overestimate summer densities, except for high-elevation regions in East Antarctica and specific areas in West Antarctica. These errors may be attributed to underestimations of short-term (or seasonal) variations in the modelled density and the limitation of RF in extrapolating values outside the training data. Overall, our study presents a potential method for estimating unknown Antarctic firn densities using known densities and satellite parameters.

Weiran Li et al.

Status: open (until 25 Oct 2023)

Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor | : Report abuse
  • RC1: 'Comment on egusphere-2023-1556', Anonymous Referee #1, 25 Sep 2023 reply

Weiran Li et al.

Weiran Li et al.


Total article views: 281 (including HTML, PDF, and XML)
HTML PDF XML Total BibTeX EndNote
204 71 6 281 4 3
  • HTML: 204
  • PDF: 71
  • XML: 6
  • Total: 281
  • BibTeX: 4
  • EndNote: 3
Views and downloads (calculated since 24 Jul 2023)
Cumulative views and downloads (calculated since 24 Jul 2023)

Viewed (geographical distribution)

Total article views: 263 (including HTML, PDF, and XML) Thereof 263 with geography defined and 0 with unknown origin.
Country # Views %
  • 1
Latest update: 30 Sep 2023
Short summary
This study used a machine learning approach to estimate the densities over the Antarctic Ice Sheet, particularly in the areas where the snow is usually dry. The motivation is to establish a link between satellite parameters to snow densities, as measurements are difficult for people to take on site. It provides valuable insights into the complexities of the relationship between satellite parameters and firn density and provides potential for further studies.