Temporal-Spatial Distribution Characteristics and Influencing Factors of Regional Agricultural Water Requirement Indicators

Abstract

The temporal-spatial distribution of agricultural water requirements constitutes an important basis for the optimal allocation of regional irrigation water resources. The primary crops (rice, maize, and soybeans) of 12 cities in Heilongjiang Province, China, were used as a case study, and reference evapotranspiration, crop water requirement, effective rainfall, irrigation requirement, and irrigation requirement index for these crops were calculated for a period of 10 years (2006–2015). The indexes of homogeneity and stability were used to reveal the characteristics of the temporal and spatial distributions based on a cloud model, i.e., a mathematical representation of qualitative concept—including expectation, Ex, entropy, En, and hyperentropy, He—that integrates the homogeneity and stability of the explored water use terms. Path analysis was used to determine the most pertinent factors influencing the irrigation requirement. The results showed a wide range in annual effective rainfall. The trends of the annual irrigation requirement and irrigation requirement index were similar, with the maximum irrigation requirement occurring in 2007. The distribution of the annual irrigation requirement was widely dispersed and that of the annual effective rainfall displayed the lowest degree of cohesion. The ranges of the various agricultural water requirement indicators between crops except for the effective rainfall were greater in the southwest than in the northeast. The distributions of the annual reference evapotranspiration, effective rainfall, irrigation requirement, and irrigation requirement index were most discrete and unstable in the Xiaoxing’anling Mountains, whereas the distributions of the annual crop water requirement, effective rainfall, and irrigation requirement index in the Sanjiang Plain were uniform and stable. Compared with the temporal distribution, the spatial distribution of each indicator was discrete and nonhomogeneous. Temperature and precipitation were the primary factors that influenced the irrigation requirement.