Spatial-temporal variability of natural factors in border fields and its effect on irrigation performance

Natural factors have been the most important parameters in the border irrigation design, such as infiltration coefficient (k), infiltration index (α), roughness (n), and slope (s₀). The irrigation performance is significantly influenced by the variability of these factors. In this study, a four-year border irrigation experiment was conducted on 11 sets of border fields, in order to reveal the spatial-temporal variability of natural factors and their impact on irrigation performance. The measured data was collected to analyze the spatial variation in the natural factors between and within the borders, as well as the interannual temporal variability of natural factors during the irrigation period. WinSRFR model was combined to simulate the impact of spatial-temporal variation in the natural factors on the flow process and irrigation performance of border irrigation. The results showed that: 1) The average spatial variation coefficients of k, α, s_0, and n of the borders were 11.00%, 4.05%, 10.28%, and 7.94%, respectively, while the average temporal variation coefficients were 26.87%, 7.73%, 4.70%, and 21.86%, respectively. Three natural factors presented the greatest impact on the irrigation performance: k, α, and n, all of which were greater temporal variability than spatial one. 2) Natural factors (k, α, and n) shared a greater impact on the flow recession process than those on the flow advance process. The natural factors presented a small impact on the flow recession time in the first half of the border, but there was a great impact on the second half of the border during the recession. Particularly, a significant impact was found at the end of the border. Therefore, the spatial-temporal variability of natural factors (k, α, and n) increased the risk of waterlogging at the end of the border. 3) The average spatial variation coefficients of k, α, and s₀ within the border were 15.98%, 7.83%, and 45.73%, respectively. There was a greater spatial variability of natural factors within the border, compared with the between borders. The variability of natural factors within the border also reduced the irrigation performance. Therefore, high irrigation performance was highly required to consider the natural factors within the border in the design of precision surface irrigation. 4) The variation of natural factors in border fields resulted in low irrigation performance, indicating the greater impact of temporal variability on irrigation performance. Therefore, it was recommended to select a typical border from farmland, in order to measure the natural factors before irrigation. And then the irrigation inflow rate and distance-based cut-off were taken as the design variables, and irrigation performance as the optimization objective. The optimal combination of irrigation inflow rate and distance-based cut-off was finally achieved after multi-objective (or single-objective) optimization. A tradeoff was obtained to balance between the excessive workload of natural factors measurement in border fields and the low irrigation performance under the natural factor variations. The findings can provide a strong reference for border decision-making on the development of precision surface irrigation.