@article{fdi:010077939,
  title = {{D}eep learning applied to glacier evolution modelling},
  author = {{B}olibar, {J}. and {R}abatel, {A}. and {G}outtevin, {I}. and {G}aliez, {C}. and {C}ondom, {T}homas and {S}auquet, {E}.},
  editor = {},
  language = {{ENG}},
  abstract = {{W}e present a novel approach to simulate and reconstruct annual glacier-wide surface mass balance ({SMB}) series based on a deep artificial neural network ({ANN}; i.e. deep learning). {T}his method has been included as the {SMB} component of an open-source regional glacier evolution model. {W}hile most glacier models tend to incorporate more and more physical processes, here we take an alternative approach by creating a parameterized model based on data science. {A}nnual glacier-wide {SMB}s can be simulated from topo-climatic predictors using either deep learning or {L}asso (least absolute shrinkage and selection operator; regularized multilinear regression), whereas the glacier geometry is updated using a glacier-specific parameterization. {W}e compare and cross-validate our nonlinear deep learning {SMB} model against other standard linear statistical methods on a dataset of 32 {F}rench {A}lpine glaciers. {D}eep learning is found to outperform linear methods, with improved explained variance (up to + 64% in space and +108% in time) and accuracy (up to +47% in space and +58% in time), resulting in an estimated r(2) of 0.77 and a root-mean-square error ({RMSE}) of 0.51 m w.e. {S}ubstantial nonlinear structures are captured by deep learning, with around 35% of nonlinear behaviour in the temporal dimension. {F}or the glacier geometry evolution, the main uncertainties come from the ice thickness data used to initialize the model. {T}hese results should encourage the use of deep learning in glacier modelling as a powerful nonlinear tool, capable of capturing the nonlinearities of the climate and glacier systems, that can serve to reconstruct or simulate {SMB} time series for individual glaciers in a whole region for past and future climates.},
  keywords = {},
  booktitle = {},
  journal = {{C}ryosphere},
  volume = {14},
  numero = {2},
  pages = {565--584},
  ISSN = {1994-0416},
  year = {2020},
  DOI = {10.5194/tc-14-565-2020},
  URL = {https://www.documentation.ird.fr/hor/fdi:010077939},
}