资源环境科技发展态势分析平台

Delineation of homogeneous reference evapotranspiration (ET0) regions is essential for different applications in hydro-meteorology. In conventional regionalization approaches, lumped (time-invariant) statistics such as mean, median or interquartile range of different hydrometeorological variables are often considered as attributes to delineate regions. Information on temporal dynamics of those variables is not utilized (as it is lost in lumped statistics), which if accounted could yield better regions. To address this, a new regionalization approach is presented in fuzzy framework in this Part 1 of a two-part series. In the proposed approach, information on temporal dynamics of predictor climate variables influencing ET0 is used for regionalization, and the delineated regions are subsequently validated for homogeneity using the predictand (ET0) related information. Effectiveness of the approach is demonstrated through a case study on India, which yielded 18 regions. They are shown to be statistically more homogeneous in ET0 when compared to the existing agro-ecological zones and regions formed using global fuzzy c-means clustering method. The homogeneous ET0 regions were found to be different from homogeneous actual evapotranspiration (ETa) regions delineated over India using the fuzzy dynamic clustering approach. Various applications of the homogeneous ET0 regions formed using the proposed approach are presented in Part 2 of this series.