@article{fdi:010076422,
  title = {{U}p-scaling of crop productivity estimations using the {A}qua{C}rop model and {GIS}-based operations},
  author = {{A}laya, {I}. and {M}asmoudi, {M}.{M}. and {J}acob, {F}r{\'e}d{\'e}ric and {B}en {M}echlia, {N}.},
  editor = {},
  language = {{ENG}},
  abstract = {{C}rop models are useful in evaluating management strategies and exploration of new practices, particularly in studies related to climate change and productivity assessment of agricultural systems. {A}t field level, biophysical crop models are generally suitable in homogeneous environments when accurate input data and calibration parameters are available. {H}owever, their use at watershed level is limited, especially in hilly areas with great variability of soils, slope, and land use. {S}ystematic method considering all terrain variabilities is time consuming since it requires high-resolution data and parameterization effort while geospatial models like {SWAT}, using simplified crop modules do not reflect the complexity of the simulated processes. {I}n this work, an alternative methodology is proposed and tested in the hilly {M}editerranean watershed of {K}amech located in the {C}ap {B}on {P}eninsula, {T}unisia ({N} 36.88 degrees, {E} 10.88 degrees); it uses the {FAO} {A}qua{C}rop biophysical model to estimate production in selected fields and scale up the results to the watershed level. {M}aps of soil, slope, and land use are combined by a {GIS} tool to obtain a database of averaged field properties and occupations. {T}hree categories of texture, depths, and slopes were considered to classify the 313 fields of the watershed into 27 soil classes and determine their respective area-weighting factor. {T}he systematic method considering all fields and the proposed method considering the 27 representative fields were used to estimate the watershed production for dominant crops: wheat, barley, and faba bean. {R}esults show a good correlation between both methods with values of relative {RMSD} in the range of 0.5-2% for biomass and 2-5% for grain yield. {D}ecile-decile analysis showed that the proposed methodology simulated almost all the observed spatial variability of yield within the watershed suggesting its suitability for productivity assessment and prediction in hilly fragmented agricultural landscape.},
  keywords = {{ZONE} {MEDITERRANEENNE} ; {TUNISIE} ; {CAP} {BON}},
  booktitle = {},
  journal = {{A}rabian {J}ournal of {G}eosciences},
  volume = {12},
  numero = {4},
  pages = {art. no 419 [11  ]},
  ISSN = {1866-7511},
  year = {2019},
  DOI = {10.1007/s12517-019-4588-5},
  URL = {https://www.documentation.ird.fr/hor/fdi:010076422},
}