王建鹏, 崔远来. 水稻灌区水量转化模型及其模拟效率分析[J]. 农业工程学报, 2011, 27(1): 22-28.
    引用本文: 王建鹏, 崔远来. 水稻灌区水量转化模型及其模拟效率分析[J]. 农业工程学报, 2011, 27(1): 22-28.
    Wang Jianpeng, Cui Yuanlai. Modified SWAT for rice-based irrigation system and its assessment[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2011, 27(1): 22-28.
    Citation: Wang Jianpeng, Cui Yuanlai. Modified SWAT for rice-based irrigation system and its assessment[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2011, 27(1): 22-28.

    水稻灌区水量转化模型及其模拟效率分析

    Modified SWAT for rice-based irrigation system and its assessment

    • 摘要: 为更合理地定量描述和研究分析南方丘陵水稻灌区水量及其转化关系,在已构建的水稻灌区分布式水量平衡模型(改进SWAT模型)的基础上进一步从模型模拟结构、稻田水量平衡要素(蒸发、降雨、渗漏和灌排)模拟和渠系渗漏影响等方面对模型予以改进完善,以期能合理体现灌区特征,充分描述灌区水文过程,提高模拟精度,为灌区水量转化和节水潜力分析研究提供有效手段。文中将完善相关改进的模型应用于漳河灌区典型流域,并将其模拟结果与原SWAT模型模拟结果比较分析。结果表明:改进SWAT模型径流模拟总量与实测值基本一致,相对误差在±10.0%之内;Nash-Suttclife效率系数达到0.57~0.76;评价系数均在0.85以上。原SWAT模型径流模拟总量仅占实测值的30%~40%;Nash-Suttclife效率系数均低于0.50,甚至出现负值;评价系数在0.76~0.91之间。改进SWAT模型水稻蒸发蒸腾量模拟值与试验值相对误差仅为-4.76%,评价系数为0.93,Nash-Suttclife效率系数高达0.85;而原SWAT模型模拟值与试验值相对误差高达-38.49%,评价系数为0.73,Nash-Suttclife效率系数为负值。可见,改进SWAT模型因全面考虑水稻灌区水文特征,其模拟效率明显优于原SWAT模型,更适合于南方丘陵区水稻灌区的水文循环模拟。

       

      Abstract: A distributed hydrological model was developed by integrating and improving the modified SWAT2000 to take quantitative research on water transformation of rice-based irrigation system in hilly area of south China. Thinking over the hydrological process and the special characteristics of rice-based irrigation system in hilly area, the main ameliorations, including the simulation framework, the factors of water balance (rainfall, ET (evportranspirtion), seepage, irrigation and drainage) and the canal seepage influence, could further improve the efficiency of simulation. Then the new modified SWAT was validated in a representative small watershed in Zhanghe Irrigation System and the simulation results were compared with the results of the original SWAT. The index of simulation efficiency, including the relative error (RE), linear regression coefficient (R2) and the Nash-Suttclife simulation efficiency coefficient (Ens), were used to the simulation results evaluation. The RE of the total runoff was within ±10.0% for the simulation values of the new modified SWAT and the observed values, but it took only 30% to 40% of the observed total runoff for the original SWAT. The R2 of the daily runoff were larger than 0.85 for the simulation values of the new modified SWAT and the observed values, while they varied from 0.76 to 0.91 for the simulation values of the original SWAT and the observed values. The Ens of the daily runoff ranged from 0.57 to 0.74 for the simulation values of the new modified SWAT and the observed values, while they were lower than 0.50 for the simulation values of the original SWAT and the observed values, even just was negative. The RE of the rice ET was only ?4.76% for the simulation values of new modified SWAT and the observed values, but it reached to -38.49% for the original SWAT and the observed values. The R2 of the rice ET reached to 0.93 for the simulation values of the new modified SWAT and the observed, while it just was 0.73 for the original SWAT and the observed values. The Ens of the rice ET was up to 0.85 for the simulation values of the new modified SWAT and observed, but it was negative for the original SWAT and the observed. The results suggested that the new modified SWAT was more effective and suited than the original SWAT. It can be a good way to study water transformation and water saving potential in rice-based irrigation system.

       

    /

    返回文章
    返回