@article{fdi:010083788,
  title = {{I}n-stream {E}scherichia coli modeling using high-temporal-resolution data with deep learning and process-based models},
  author = {{A}bbas, {A}. and {B}aek, {S}. and {S}ilvera, {N}orbert and {S}oulileuth, {B}. and {P}achepsky, {Y}. and {R}ibolzi, {O}livier and {B}oithias, {L}. and {C}ho, {K}. {H}.},
  editor = {},
  language = {{ENG}},
  abstract = {{C}ontamination of surface waters with microbiological pollutants is a major concern to public health. {A}lthough long-term and high-frequency {E}scherichia coli ({E}. coli) monitoring can help prevent diseases from fecal pathogenic microorganisms, such monitoring is time-consuming and expensive. {P}rocess-driven models are an alternative means for estimating concentrations of fecal pathogens. {H}owever, process-based modeling still has limitations in improving the model accuracy because of the complexity of relationships among hydrological and environmental variables. {W}ith the rise of data availability and computation power, the use of data-driven models is increasing. {I}n this study, we simulated fate and transport of {E}. coli in a 0.6 km(2) tropical headwater catchment located in the {L}ao {P}eople's {D}emocratic {R}epublic ({L}ao {PDR}) using a deep-learning model and a process-based model. {T}he deep learning model was built using the long short-term memory ({LSTM}) methodology, whereas the process-based model was constructed using the {H}ydrological {S}imulation {P}rogram-{FORTRAN} ({HSPF}). {F}irst, we calibrated both models for surface as well as for subsurface flow. {T}hen, we simulated the {E}. coli transport with 6 min time steps with both the {HSPF} and {LSTM} models. {T}he {LSTM} provided accurate results for surface and subsurface flow with 0.51 and 0.64 of the {N}ash-{S}utcliffe efficiency ({NSE}) values, respectively. {I}n contrast, the {NSE} values yielded by the {HSPF} were -0.7 and 0.59 for surface and subsurface flow. {T}he simulated {E}. coli concentrations from {LSTM} provided the {NSE} of 0.35, whereas the {HSPF} gave an unacceptable performance with an {NSE} value of -3.01 due to the limitations of {HSPF} in capturing the dynamics of {E}. coli with land-use change. {T}he simulated {E}. coli concentration showed the rise and drop patterns corresponding to annual changes in land use. {T}his study showcases the application of deep-learning-based models as an efficient alternative to process-based models for {E}. coli fate and transport simulation at the catchment scale.},
  keywords = {{LAOS}},
  booktitle = {},
  journal = {{H}ydrology and {E}arth {S}ystem {S}ciences},
  volume = {25},
  numero = {12},
  pages = {6185--6202},
  ISSN = {1027-5606},
  year = {2021},
  DOI = {10.5194/hess-25-6185-2021},
  URL = {https://www.documentation.ird.fr/hor/fdi:010083788},
}