Kohonen Self-Organizing Map as a software sensor estimator of reference crop evapotranspiration

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    Abstract

    Reference crop evapotranspiration (ETo) estimation is of importance in irrigation water management for the calculation of crop water requirements and its scheduling, rainfall runoff modelling and numerous other water resources studies. This paper developed the Kohonen Self-Organizing Map (KSOM), unsupervised artificial neural networks software sensors to predict the ETo. This was achieved by using the powerful clustering capability of the KSOM to analyse the multi-dimensional data array of estimated Penman-Monteith ETo and different subsets of input climatic variables. Data obtained at two locations, Edinburgh (UK) and Udaipur (India), were used in order to demonstrate the versatility of the approach. The findings indicate that the KSOM-based ETo estimates were in good agreement with those obtained using the conventional FAO Penman-Monteith formulation at both locations. A further comparison of the KSOM estimates with other commonly used empirical methods for the ETo, e.g. Thornthwaite, Priestley-Taylor and Hargreaves, showed that the former were far superior. This offers significant potential for accurate estimation of the ETo in regions of the world where the needed climatic data are unavailable for the implementation of the full Penman-Monteith formulation.

    Original languageEnglish
    Title of host publicationHydro-climatology: Variability and Change
    Subtitle of host publicationProceedings of the Symposium JHS02 held during the IUGG GA in Melbourne (28 June-7 July 2011)
    EditorsSW Franks, E Boegh, E Blyth, DM Hannah, KK Yilmaz
    Place of PublicationWALLINGFORD
    PublisherIAHS Press
    Pages7-14
    Number of pages8
    Volume344
    ISBN (Print)978-1-907161-19-3
    Publication statusPublished - 2011
    Event25th General Assembly of the International Union of Geodesy and Geophysics - Melbourne
    Duration: 28 Jun 20117 Jul 2011

    Conference

    Conference25th General Assembly of the International Union of Geodesy and Geophysics
    CityMelbourne
    Period28/06/117/07/11

    Fingerprint Dive into the research topics of 'Kohonen Self-Organizing Map as a software sensor estimator of reference crop evapotranspiration'. Together they form a unique fingerprint.

    Cite this