| 引用本文: | 侯靳锦,孙筱璐,王碧胜,等.不同水分条件下APEX模型参数的敏感性评价和不确定性分析[J].灌溉排水学报,0,():-. |
| HOU Jinjin,SUN Xiaolu,WANG Bisheng,et al.不同水分条件下APEX模型参数的敏感性评价和不确定性分析[J].灌溉排水学报,0,():-. |
|
| 摘要: |
| 【目的】农业系统模型参数敏感性分析是模型验证和应用的关键，为研究不同土壤、气候、管理因素对模型参数敏感性的影响，【方法】本研究基于胶东地区冬小麦水分控制试验（2016—2019年），采用Sobol法、Morris法和FAST法分析评价了不同水分条件下APEX（Agricultural Policy/Environmental eXtender）模型参数的敏感性，探讨了不同灌溉管理、降水年型及地下水位对模型参数敏感性的影响。【结果】在当地试验条件下（地下水位为1.25 m），影响冬小麦生物量、产量及农田蒸散的最敏感参数均为RDMX（最大根深度），影响叶面积指数的最敏感的参数是DMLA（最大潜在叶面积指数）；当地下水位设为5 m时，影响农田蒸散、作物产量和生物量的敏感性参数变为PARM38（水分胁迫计算权重系数）和RWPC1（萌发期根系生物量占比）。随灌水量的增加，RDMX参数敏感性总体降低，而DMLA、DLAI（叶面积生长达到最高点的生育时期）和WA（潜在光能利用率）的敏感性总体升高，RDMX在干旱年份敏感性指数明显高于湿润年份。模型不确定性分析表明：冬小麦生物量、产量和农田蒸散观测值均在模拟结果的5%~95%置信区间之内，模拟结果具有较高的可信度。【结论】三种分析方法筛选的最敏感参数具有较好的一致性，但其敏感性在不同灌溉、降水年型或地下水位条件下差异明显，考虑到计算效率和准确性，Morris法较适用于APEX模型参数敏感性分析，研究结果为APEX模型在不同水分条件下参数优化和应用提供了重要依据。 |
| 关键词: APEX模型；敏感性分析；水分胁迫；Sobol法；Morris法；FAST法；不确定性分析 |
| DOI： |
| 分类号:S512.1 |
| 基金项目:国家自然科学(31671627)；山东省自然科学基金青年项目(ZR2021QC113)；青岛农业大学博士启动基金(6631118021) |
|
| Sensitivity and uncertainty analysis of APEX model parameters under different soil moisture conditions |
|
HOU Jinjin, SUN Xiaolu, WANG Bisheng, YANG Xiaohui, XU Mengjie, FANG Quanxiao
|
|
Qingdao Agricultural University College of agriculture
|
| Abstract: |
| 【Background】 Sensitivity analysis of agricultural system model is critical to model parameterization and application.【Objective】 In order to investigate the effects of different soil, climate, management factors on the sensitivity of model parameter under various soil water conditions.【Method】Based on the irrigation experiment of Winter Wheat at Jiaodong area from 2016 to 2019, the Sobol, Morris and FAST methods were used to analyze the sensitivities of parameters that closely related to crop growth and water stress in APEX (Agricultural Policy/Environmental eXtender) model as influenced by groundwater levels, rainfall pattern and irrigation management. 【Result】Under local experimental conditions (groundwater table with 1.25 m), the most sensitive parameter affecting evapotranspiration, biomass and yield was RDMX (maximum root depth), and the most sensitive parameter affecting leaf area index (LAI) was DMLA (maximum potential leaf area index). When the groundwater table was set at 5 m, the sensitive parameters affecting crop evapotranspiration and yield were obviously different, and the parameters of PARM38 (weight coefficient of water stress calculation) and RWPC1 (proportion of root biomass during germination), instead of RDMX under 1.25 m groundwater table, became the most sensitive parameters. All the three methods indicated that, with the increase of irrigation water, the sensitivity of RDMX decreased, but the sensitivities of DMLA, DLAI (peak point in growth season) and WA (potential light energy utilization) increased. The sensitivity indexes of RDMX in dry years were significantly higher than that in humid years, while the sensitivity of DMLA showed the opposite trend. The uncertainty analysis showed that the observed wheat biomass, yield and evapotranspiration were all within the 5%~95% confidence interval of the simulated data, indicating a high reliability of the model in the region. 【Conclusion】The most sensitive parameters screened by the Sobol method, Morris method and FAST method showed good consistency, but their sensitivity indexes were obvious different among the irrigation treatments, rainfall patterns and groundwater table levels. Considering the calculation efficiency and accuracy, Morris method is more suitable for parameter sensitivity analysis of APEX model. These results provided useful information on APEX calibration and application under various soil water conditions. |
| Key words: APEX model; sensitivity analysis; water stress; Sobol; Morris; Fast; uncertainty analysis |
