利用联合多重分形方法识别在不同尺度上对土壤水分特征曲线空间变异性都具有显著影响的因素,基于得出的结论建立考虑尺度效应的小尺度和大尺度van Genuchten模型参数的土壤传递函数,探讨了将小尺度上得出的联合多重分形结论进行尺度扩展,应用到大尺度上的可行性。结果表明:小尺度和大尺度上基于0~20 cm和20~40 cm土层van Genuchten模型参数土壤传递函数预测的土壤含水率均方根误差分别为0.038 6、0.047 3和0.027 0、0.030 4,可将小尺度上进行联合多重分形分析得出的结论进行尺度扩展,应用到大尺度上,且建立的土壤传递函数具有较强的理论基础和较高的预测精度。
The factors that have obvious effects on spatial variability at the different scales were identified with joint multifractal, and pedotransfer functions of van Genuchten model parameters which considered scale effect at small and large scale were established based on the above educed conclusion, furthermore, feasibility that concluded and educed at small scale with joint multifractal and applied at large scale was explored. The results showed that root mean square error of forecasted soil water content based on pedotransfer functions of van Genuchten model parameters in 0 - 20 cm and 20 - 40 cm soil layers at small scale was 0. 038 6 and 0. 047 3 respectively, and RMSE of forecasted soil water content based on pedotransfer functions of van Genuchten model parameters in 0 - 20 cm and 20- 40 cm soil layers at large scale was 0. 027 0 and 0. 030 4 respectively; conclusions educed at small scale with joint multifractal could be applied at large scale, and pedotransfer functions based on these conclusions had strong theoretical foundation and high precision.