FACILITATION OF OPERATION AND MAINTENANCE ACTIVITIES OF IRRIGATION NETWORKS USING A K‐MEANS CLUSTERING METHOD: CASE STUDY OF THE GHAZVIN IRRIGATION NETWORK

Abstract

ABSTRACTOne of the most important factors in saving limited water resources is improving the performance of irrigation canal networks. To achieve this purpose, improving the operation and maintenance (O&M) activities of irrigation networks is necessary. Most current O&M approaches do not consider the spatial differentiation of networks. Detection of homogeneous areas provides managers with guidelines for planning O&M activities and allocating budgets, labour, and machinery resources. In similar irrigation areas, operators exchange their experience more efficiently. In this paper, the application of K‐means clustering algorithms as a quantitative benchmark for exploring homogeneous areas with similar physical attributes outside the irrigation network region is presented. A K‐means clustering algorithm is applied for spatial clustering of the Ghazvin irrigation network. Five physical attributes of canal reaches, viz. length, capacity, number of off‐takes, number of conveyance structures and the irrigated area covered, are considered. Using a clustering validity index, the 162 canal reaches in the Ghazvin irrigation network are clustered in 10 groups, with members varying from 5 to 30 reaches. Regionalization and assigning the groups to colonies out of the irrigation network district provide a context for better and easier decision making by managers. Copyright © 2011 John Wiley & Sons, Ltd.