@article{fdi:010080609,
  title = {{A} {PCA} spatial pattern based artificial neural network downscaling model for urban flood hazard assessment},
  author = {{C}arreau, {J}ulie and {G}uinot, {V}.},
  editor = {},
  language = {{ENG}},
  abstract = {{W}e present two statistical models for downscaling flood hazard indicators derived from upscaled shallow water simulations. {T}hese downscaling models are based on the decomposition of hazard indicators into linear combinations of spatial patterns obtained from a {P}rincipal {C}omponent {A}nalysis ({PCA}). {A}rtificial {N}eural {N}etworks ({ANN}s) are used to model the relationship between low resolution ({LR}) and high resolution ({HR}) information drawn from hazard indicators. {I}n both statistical models, the {PCA} features, i.e. the linear weights of the spatial patterns, of the {LR} hazard indicator are taken as inputs to the {ANN}s. {I}n the first model, there is one {ANN} per {HR} cell where the hazard indicator is to be estimated and the output of the {ANN} is the hazard indicator value at that cell. {I}n the second model, there is a single {ANN} for the whole {HR} mesh whose outputs are the {PCA} features of the {HR} hazard indicator. {A}n estimate of the hazard indicator is obtained by combining the {ANN}'s outputs with the {HR} spatial patterns. {T}he two statistical downscaling models are evaluated and compared at estimating the water depth and the norm of the unit discharge, two common hazard indicators, on simulations from five synthetic urban configurations and one field-test case. {A}nalyses are carried out in terms of relative absolute errors of the statistical downscaling model with respect to the {LR} hazard indicator. {T}hey show that (i) both statistical downscaling models provide estimates that are more accurate than the {LR} hazard indicator in most cases and (ii) the second downscaling model yields consistently lower errors for both hazard indicators for all flow scenarios on all configurations considered. {T}he statistical models are three orders of magnitude faster than {HR} flow simulations. {U}sed in conjunction with upscaled flood models such as porosity models, they appear as a promising operational alternative to direct flood hazard assessment from {HR} flow simulations.},
  keywords = {{S}hallow water models ; {P}orosity models ; {F}low variables for hazard assessment ; {M}ultisite statistical downscaling ; {A}rtificial neural networks ; {P}rincipal component analysis},
  booktitle = {},
  journal = {{A}dvances in {W}ater {R}esources},
  volume = {147},
  numero = {},
  pages = {103821 [15 ]},
  ISSN = {0309-1708},
  year = {2021},
  DOI = {10.1016/j.advwatres.2020.103821},
  URL = {https://www.documentation.ird.fr/hor/fdi:010080609},
}