南方红壤区植被覆盖因子估算模型构建与验证

Model construction and verification of vegetation cover and management factor in southern red soil region of China

研究红壤区植被覆盖因子变化规律、构建植被覆盖因子计算模型可为该区域水土流失防治及动态监测提供科学依据。该研究基于广东省五华县水土保持试验推广站200余场降雨试验观测数据,分析了影响水土流失的各关键因子与次降雨土壤侵蚀模数(Individual Rainfall Event Soil Erosion Modulus,ISEM)间定量关系,基于中国土壤流失方程计算了系列次降雨事件不同植被覆盖度对应植被覆盖因子值,构建了植被覆盖因子值与植被覆盖度间数学模型,从点、面2个尺度对模型计算精度进行了验证。结果表明:1)ISEM随植被覆盖度增加而降低;ISEM与雨前土壤表层含水率间存在显著正相关对数关系;ISEM随坡度增加呈先增后减的变化规律;ISEM随次降雨量、次降雨侵蚀力和次降雨径流深呈显著正相关线性变化关系。2)构建了植被覆盖因子值与植被覆盖度间二阶指数衰减模型,该模型决定系数和纳什系数分别达0.947和0.876。点尺度验证结果表明90%样本模型计算值与观测值相对误差均小于0.30;面尺度验证结果表明,70%~80%的植被覆盖因子计算值相对误差不超过0.1。总体而言,该模型计算精度较为理想,但由于研究对象典型的时空尺度特征,仍需要更多观测数据对该模型进行完善和验证。研究成果可为深入理解红壤区土壤侵蚀规律、水土流失动态监测提供有益参考。

Studying the change characteristic of vegetation cover factor in the red soil region and constructing vegetation cover factor calculation model can provide valuable information for regional soil erosion control and dynamic monitoring.Based on more than 200 field rainfall events in Soil and Water Conservation Experimental Extension Station of Wuhua County in Guangdong province,this study analyzed the quantitative relationship between the soil erosion modulus and factors which affected soil erosion.The plants were mainly Pinus massoniana Lamb.,Eucalyptus gran dis Hill,Citrus maxima(Burm.)Merr.,Schima kwangtungensis Hung T.Chang,Rosa sericea Lindl.,Dicranopteris dichoto ma(Thunb.)Bernh.,Eriachne pallescens R.Br.and so on.The soil field water holding capacity was 26.7%-36.7%.The soil organic matter was 078%-1.69%.The artificial rainfall simulation device was used to produce rainfall.The effective rainfall area was 2 m by 3 m.The rainfall intensity was 0.5-2.5 mm/min.The rainfall dimeter was 0.6-2.2 mm.The rainfall uniformity was 81.5%-93.6%.This study obtained the vegetation cover factor values under different underlying surfaces based on the Chinese Soil Loss Equation,and constructed the quantitative model between the vegetation cover factor value and vegetation cover.The model was established based on data from 98 rainfall tests in 2016 and 2018,from 81 natural rainfall events in 2016 and from 29 natural rainfall events in 2017.Furthermore,the tested data were used at two spatial scales to verify the calculation accuracy of the model.The verification data was from 33 natural rainfall events in 2019 at Wubeihe station.In addition,data from Wuhua county was also used for model verification.The following key results were obtained:1)The negative power function could be used to describe the relationship between the soil erosion modulus and vegetation cover,and the fitted equation reached a significant level(P=1.11×10-6).There was a significant positive logarithmic function relation between the soil erosion modulus and the soil moisture at 0-15 cm soil depth before each rainfall event(P=9.34×10-16).The soil erosion modulus increased firstly and then decreased with the increase of slope gradient.The soil erosion modulus increased significantly with the increase of the individual rainfall amount,the individual rainfall erosivity and the runoff depth(P≤1.08×10-13).2)Based on the comparison of the statistics of fitted equations,the second-order exponential decay model was determined to quantitatively characterize the dynamic relationship between vegetation cover factor and vegetation cover.The fitted equation between vegetation cover factor and vegetation cover based on the second-order exponential decay model reached to a significant level(P=0.003),and the coefficient of determination and Sutcliffe efficiency coefficient were 0.947 and 0.876.The point scale verification result showed that the relative error of 90%verified samples were less than 0.30.The regional scale verification results showed that the absolute relative error of 70%-80%verified samples was less than 0.1.In summary,the calculation accuracy of the model was relatively ideal.The research results above could provide a useful method for in-depth understanding of the soil erosion characteristic and dynamic monitoring in red soil regions.