@article{fdi:010069238,
  title = {{A} comparison of multiple regression and neural network techniques for mapping in situ p{CO}2 data},
  author = {{L}ef{\`e}vre, {N}athalie and {W}atson, {A}. {J}. and {W}atson, {A}. {R}},
  editor = {},
  language = {{ENG}},
  abstract = {{U}sing about 138 000 measurements of surface p{CO}2 in the {A}tlantic subpolar gyre (50–70°{N}, 60–10°{W}) during 1995–1997, we compare two methods of interpolation in space and time: a monthly distribution of surface {CO}2 constructed using multiple linear regressions on position and temperature, and a self-organizing neural network approach. {B}oth methods confirm characteristics of the region found in previous work, i.e. the subpolar gyre is a sink for atmospheric {CO}2 throughout the year, and exhibits a strong seasonal variability with the highest undersaturations occurring in spring and summer due to biological activity. {A}s an annual average the surface p{CO}2 is higher than estimates based on available syntheses of surface p{CO}2. {T}his supports earlier suggestions that the sink of {CO}2 in the {A}tlantic subpolar gyre has decreased over the last decade instead of increasing as previously assumed. {T}he neural network is able to capture a more complex distribution than can be well represented by linear regressions, but both techniques agree relatively well on the average values of p{CO}2 and derived fluxes. {H}owever, when both techniques are used with a subset of the data, the neural network predicts the remaining data to a much better accuracy than the regressions, with a residual standard deviation ranging from 3 to 11 µatm. {T}he subpolar gyre is a net sink of {CO}2 of 0.13 {G}t-{C} yr−1 using the multiple linear regressions and 0.15 {G}t-{C} yr−1 using the neural network, on average between 1995 and 1997. {B}oth calculations were made with the {NCEP} monthly wind speeds converted to 10 m height and averaged between 1995 and 1997, and using the gas exchange coefficient of {W}anninkhof.},
  keywords = {{ATLANTIQUE}},
  booktitle = {},
  journal = {{T}ellus {S}eries {B} : {C}hemical and {P}hysical {M}eteorology},
  volume = {57},
  numero = {5},
  pages = {375--384},
  ISSN = {0280-6509},
  year = {2005},
  DOI = {10.3402/tellusb.v57i5.16565},
  URL = {https://www.documentation.ird.fr/hor/fdi:010069238},
}