@article{fdi:010073081,
  title = {{M}apping ecosystem services at the regional scale : the validity of an upscaling approach},
  author = {{L}e {C}lec'h, {S}. and {S}loan, {S}. and {G}ond, {V}. and {C}ornu, {G}. and {D}ecaens, {T}. and {D}ufour, {S}. and {G}rimaldi, {M}ichel and {O}szwald, {J}.},
  editor = {},
  language = {{ENG}},
  abstract = {{M}apping ecosystem services ({ES}) over large scales is important for environmental monitoring but is often prohibitively expensive and difficult. {W}e test a hybrid, low-cost method of mapping {ES} indicators over large scales in {P}ara {S}tate, {B}razil. {F}our {ES} indicators (vegetation carbon stocks, biodiversity index, soil chemical quality index and rates of water infiltration into soil) were measured in the field and then summarized spatially for regional land-cover classes derived from satellite imagery. {T}he regionally mapped {ES} values correlated strongly with independent and local measures of {ES}. {F}or example, regional estimates of the vegetation carbon stocks are strongly correlated with actual measures derived from field samples and validation data (significant anova test - p-value=4.51(e-9)) and differed on average by only 20{M}g/ha from the field data. {O}ur spatially-nested approach provides reliable and accurate maps of {ES} at both local and regional scales. {L}ocal maps account for the specificities of an area while regional maps provide an accurate generalization of an {ES}' state. {S}uch up-scaling methods infuse large-scale {ES} maps with localized data and enable the estimation of uncertainty of at regional scales. {O}ur approach is first step towards the spatial characterization of {ES} at large and potentially global scales.},
  keywords = {{E}cosystem services mapping ; up-scaling ; remote sensing ; soil chemical ; quality ; soil infiltrability ; biodiversity ; carbon stocks ; {B}razilian ; {A}mazon ; {BRESIL} ; {AMAZONE}},
  booktitle = {},
  journal = {{I}nternational {J}ournal of {G}eographical {I}nformation {S}cience},
  volume = {32},
  numero = {8},
  pages = {1593--1610},
  ISSN = {1365-8816},
  year = {2018},
  DOI = {10.1080/13658816.2018.1445256},
  URL = {https://www.documentation.ird.fr/hor/fdi:010073081},
}