Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Master → Pays Bas → A spatial analysis of water productivity in an irrigated rice scheme using remote sensing and ground-based data : a case study of the Office du Niger irrigation scheme in Mali

Wageningen University (2011)

A spatial analysis of water productivity in an irrigated rice scheme using remote sensing and ground-based data : a case study of the Office du Niger irrigation scheme in Mali

Dutrieux, L.P.

Titre : A spatial analysis of water productivity in an irrigated rice scheme using remote sensing and ground-based data : a case study of the Office du Niger irrigation scheme in Mali

Auteur : Dutrieux, L.P.

Université de soutenance  : Wageningen University

Grade : Master of Science (MS) 2011

Résumé
In the context of growing pressure on water resources, due to the increasing world population and its dietary changes, increasing agricultural water productivity appears as a necessity for a sustainable water resources use in the future. In this regard, a remote sensing study was carried out in 2006 to assess the irrigation performance of the Office du Niger, a large ricebased system in Mali. The results were published in (Zwart and Leclert, 2010). It was revealed that large differences exist at various scales within the system when the estimated rice yields, evapotranspiration and crop water productivity were considered. It was acknowledged that the limitations concerned the fixed growing period and the fixed harvest index that were used. Moreover, the variations were assessed, but not yet explained. The purpose of this study, presented as a follow up of the work of Zwart and Leclert (2010), is twofold. First of all, it describes the improvements in terms of growing season and harvest index that were made to the maps of seasonal yields, evapotranspiration and crop water productivity. Secondly, the spatial variations of water productivity were explained by relating the enhanced maps to different datasets of the bio-physical environment. Measurements taken from 24 farmers’ fields show that, with an average value of 0.54 (grain moisture corrected to 14%), harvest index has a fairly limited variability in the conditions of the Office du Niger (Cv = 0.05). It was concluded, given its stability that the use of one single harvest index value is justified in remote sensing studies. Further improvements were achieved concerning the water productivity map with the introduction of variable growing periods, defined from the NDVI time profiles on a pixel basis. These growing periods were thereafter considered in the integration of biomass production and evapotranspiration calculated with the Surface Energy Balance Algorithm for Land (SEBAL) model. Concerning the explanation of spatial variations, planting period was found to be the most influential parameter on water productivity, whereas no relationships could be found between water productivity and soil or water-related variables. It is discussed that the scale of analysis, and the fact that the study only considers one year, may have reduced the chance of finding relationships between water productivity and secondary data and could not allow the validation of the results found. Recommendations are proposed, including the need for more research, considering several years and different scales of analysis.

Mots clés : rice / irrigation / productivity / water use efficiency / spatial analysis / remote sensing / mali

Présentation

Version intégrale (2 Mb)

Page publiée le 13 février 2015, mise à jour le 13 octobre 2018