| 引用本文: | 高晓飞,高燕,刘刚.使用有限数据推求土壤水分特征曲线VG模型的参数[J].灌溉排水学报,0,():-. |
| GAO Xiaofei,GAO Yan,LIU Gang.使用有限数据推求土壤水分特征曲线VG模型的参数[J].灌溉排水学报,0,():-. |
|
| |
|
|
| 本文已被：浏览 37次 下载 0次 |
|
|
|
| 使用有限数据推求土壤水分特征曲线VG模型的参数 |
|
高晓飞1, 高燕2, 刘刚1
|
|
1.北京师范大学 地表过程与资源生态国家重点实验室;2.松辽水利委员会 松辽流域水土保持监测中心站
|
|
| 摘要: |
| 【目的】使用一种简易方法和有限数据来获取土壤水分特征曲线VG模型的参数。【方法】选择了黑土、褐土、黄绵土、红壤和紫色土5种地带性土壤,实测了土壤饱和含水量(θs)、0.33bar土壤含水量(θ0.33bar)、最大吸湿水(θ97%RH)和空气相对湿度为43%时的土壤吸湿水(θ43%RH),并以θ43%RH为残余含水量(θr),来推导VG模型的参数。将推导出的参数带入到VG模型中,进行了土壤水分特征曲线的预报,并与实测的土壤含水量进行了比较。【结果】结果发现,该方法能较好的模拟褐土的土壤水分特征曲线,模拟值和实测值的残差一般在0.01cm3.cm-3以内；也可以较好的模拟黄绵土和红壤在0.33-15bar的土壤含水量,残差一般在±0.01cm3.cm-3以内；但黑土的模拟结果相对较差,在低于0.33bar时,残差为负,在-0.03 cm3.cm-3以内,而在0.33-15bar,残差为正,达到0.04 cm3.cm-3；紫色土的模拟结果最差,在低于0.33bar时,残差为负,达到了-0.07 cm3.cm-3；但在0.33-15bar,情况有所改善,残差在0.03 cm3.cm-3以内。通过配对t检验,认为模拟值和实测值无显著差异,二者的相关系数(r)为0.990。造成模拟值和实测值差异的原因可能是采样时紫色土密度较低,土壤中有较多的大孔隙,而黑土可能因有机质含量较高具有较好的结构性导致持水能力较强。【结论】采用土壤饱和含水量(θs)、0.33bar土壤含水量(θ0.33bar)、最大吸湿水(θ97%RH)和空气相对湿度为43%时的土壤吸湿水(θ43%RH)推导的VG模型参数,可以较好地预报土壤水分特征曲线。但对土壤密度较低、存在大孔隙的土壤,及有机质含量较高或结构性较好的土壤,可能会导致部分模拟值和实测值产生较明显的差异。前者可能会影响低吸力段的模拟结果(<0.33bar),后者则可能影响土壤有效含水量段的模拟结果(0.33-15bar)。 |
| 关键词: VG模型;土壤吸湿水;土壤水分特征曲线参数;预报残差 |
| DOI： |
| 分类号:S152.7 |
| 基金项目:多尺度土壤复合侵蚀预测模型及耕地质量退化风险（2021YFD1500704） |
|
| Using Limited Data to Derive VG Model Parameters for Soil Water Characteristic Curves |
|
GAO Xiaofei1, GAO Yan2, LIU Gang1
|
|
1.State Key Laboratory of Earth Surface Processes and Resource Ecology,Beijing Normal University;2.Water and Soil Conservation Monitoring Center of Songliao River Basin,Songliao Water Resources Commission
|
| Abstract: |
| 【Background】Soil-water characteristic curves(SWCC) is an important input parameter in water and solute transport modeling of unsaturated zone. The VG (van Genuchten) model is a good method to simulate the SWCC. Direct measurements of hydraulic parameters are so tedious and time-consuming that many methods were explored to acquire the VG model parameters in past several decades. But, these methods need sufficient data, too. 【Objective】The objective of this research is to develop a simple method using limited data to obtain the parameters of the VG model.【Method】Five regional soils were selected for the experiment, including black soil, cinnamon soil, loessial soil, red earth, and purple soil. Their saturated soil moisture (θs), soil moisture under 0.33 bar pressure (θ0.33bar), hygroscopicity at 97% relative humidity of air (θ97%RH) and 43% relative humidity of air (θ43%RH) were measured. Then, the parameters of the VG model were derived from θs, θ0.33bar, θ97%RH and by using the θ43%RH as residual moisture (θr). The derived parameters were taken into the VG model, and the newly obtained SWCC was compared with the measured soil moisture. 【Result】It was found that this method could simulate the SWCC of cinnamon soil well, the residual between the simulated and measured values is generally within 0.01cm3. cm-3; It could also simulate the soil moisture of loessial soil and red earth at 0.33-15bar well, with residuals generally within ± 0.01 cm3. cm-3. However, the simulation results of black soil are relatively poor, the residual is negative, within -0.03 cm3. cm-3 in low suction (< 0.33 bar) section, while at 0.33-15 bar section, the residual is positive, reaching 0.04 cm3. cm-3. The simulation results of purple soil are the worst, with a negative residual -0.07 cm3. cm-3 in low suction (<0.33 bar); But at 0.33-15 bar, the situation improved and the residual was within 0.03 cm3. cm-3. Overall, there is no significant difference between the simulated and measured value through paired t-test, and the correlation coefficient between them is 0.990. The poor results in purple soil could be derived from a low density and large pores in the soil sample, while in black soil may be derived from a strong water holding capacity due to its high organic matter content and good soil structure. 【Conclusion】The VG model parameters derived by using θs, θ0.33bar, θ97%RH and θ43%RH, can effectively predict SWCC. However, for soils with low soil density and large pores, or soils with high organic matter content or good soil structure, it may lead to significant differences in some simulated values. The former may affect the simulation results of the low suction section (<0.33 bar), while the latter may affect the simulation results of the available soil moisture section (0.33-15 bar). |
| Key words: VG model; soil hygroscopicity; parameters of SWCC; forecast residuals |
|
|
|
|
