@article{fdi:010091909,
  title = {{S}oil salinity mapping of plowed agriculture lands combining radar {S}entinel-1 and optical {S}entinel-2 with topographic data in machine learning models},
  author = {{T}ola, {D}. and {S}atg{\'e}, {F}r{\'e}d{\'e}ric and {Z}olá, {R}. {P}. and {S}ainz, {H}. and {C}ondori, {B}. and {M}iranda, {R}. and {Y}ujra, {E}. and {M}olina-{C}arpio, {J}. and {H}ostache, {R}enaud and {E}spinoza-{V}illar, {R}.},
  editor = {},
  language = {{ENG}},
  abstract = {{T}his study assesses the relative performance of {S}entinel-1 and -2 and their combination with topographic information for plow agricultural land soil salinity mapping. {A} learning database made of 255 soil samples' electrical conductivity ({EC}) along with corresponding radar ({R}), optical ({O}), and topographic ({T}) information derived from {S}entinel-2 ({S}2), {S}entinel-1 ({S}1), and the {SRTM} digital elevation model, respectively, was used to train four machine learning models ({D}ecision tree-{DT}, {R}andom {F}orest-{RF}, {G}radient {B}oosting-{GB}, {E}xtreme {G}radient {B}oosting-{XGB}). {E}ach model was separately trained/validated for four scenarios based on four combinations of {R}, {O}, and {T} ({R}, {O}, {R}+{O}, {R}+{O}+{T}), with and without feature selection. {T}he {R}ecursive {F}eature {E}limination with k-fold cross validation ({RFE}cv 10-fold) and the {V}ariance {I}nflation {F}actor ({VIF}) were used for the feature selection process to minimize multicollinearity by selecting the most relevant features. {T}he most reliable salinity estimates are obtained for the {R}+{O}+{T} scenario, considering the feature selection process, with {R}2 of 0.73, 0.74, 0.75, and 0.76 for {DT}, {GB}, {RF}, and {XGB}, respectively. {C}onversely, models based on {R} information led to unreliable soil salinity estimates due to the saturation of the {C}-band signal in plowed lands.},
  keywords = {soil salinity mapping ; plowed lands ; machine learning ; {S}entinel-1 ; {S}entinel-2 ; {BOLIVIE} ; {ANDES}},
  booktitle = {},
  journal = {{R}emote {S}ensing},
  volume = {16},
  numero = {18},
  pages = {3456 [18 ]},
  year = {2024},
  DOI = {10.3390/rs16183456},
  URL = {https://www.documentation.ird.fr/hor/fdi:010091909},
}