Studies On The Infiltration Water Storage And Prevention Of Autumn Crops In Arid Loess Sloping Fields

Vegetation is always the protector for surface coverage of terrestrial ecosystem.On the formation process of soil erosion,many studies have been conducted to demonstrate that vegetation can intercept rainfall,increase soil infiltration rates and directly affect the intensity and frequency of surface runoff and soil erosion,and surface runoff and soil erosion has decreases with the increase in vegetation coverage to a large extent.There are essential differences between sloping farmland and forest land,shrub land and grass land.The sloping farmland is highly frequently cultivated without litter layers,and the condition of sloping farmland is similar to the bare land at seeding stage,so the sloping farmland of crops has its unique characteristics on affecting soil erosion.However,there are little studies on these characteristics of sloping farmland.For this reason,this study mainly focused on the sloping farmland with autumn crops of millet,mazie and soybean cultivated in Loess Plateau,and studied the effects of crops on rainfall infiltration and soil erosion prevention through indoor and outdoor artificial rainfall stimulation combining the theory and technology of math,soil science and computer science.The main results as following:1.In the sloping farmland,millet,mazie and soybean all had significant effects on runoff and sediment yield from sloping farmland.The capacity of effects were not only related to the growth stages of crops and leaf area index,but also connected with the rainfall intensity and slope gradient.Under the same experimental conditions,effects of crops at the seeding stage were the least and would increase at later stages.When the leaf area index of millet was 0.78,the mean runoff and sediment rate was 32.64 L/m² and 167.65 g/m²,which was 57.88% and 16.96% of that on the bare land,respectively.The mean runoff and sediment rate was 19.86 L/m² and 38.65 g/m² when the leaf area of millet was 3.05,which was 57.88% and 16.96% of that on the bare land,respectively.The mean runoff and sediment rate was 29.30 L/m² and 164.61 g/m² when the leaf area index of mazie was 1.24,which was 87.88% and 69.16% of that on the bare land,respectively.when the leaf area of mazie was 4.30,the mean runoff and sediment rate was 19.97 L/m² and 60.59 g/m²,which was 57.88% and 16.96% of that on the bare land,respectively.When the leaf area index of soybean was 1.85,the mean runoff and sediment rate was 30.18 L/m² and 163.91 g/m²,which was 88.58% and 74.48% of that on the bare land,respectively;when the leaf area of soybean was 4.30,the mean runoff and sediment rate was 2 1.14L/m² and 51.46 g/m²,which was 62.05% and 23.38% of that on the bare land,respectively.The prediction model of comprehensive influencing factors of runoff and sediment can be established through the particle swarm intelligence algorithms.2.Millet,mazie and soybean all perform well in soil and water conservation.Under the experiment conditions,the maximum,minimum and mean value of water conservation benefits of millet was 53.7%,4.6% and 28.1%,respectively,and 92.8%,24.8% and 65.1% for soil conservation benefits,respectively;the maximum,minimum and mean value of water conservation benefits of mazie was 49.7%,4.9% and 26.9%,respectively,and 87.1%,22.4% and 55.4% for soil conservation benefits,respectively;the maximum,minimum and mean value of water conservation benefits of soybean was 51.3%,5.2% and 24.0%,respectively,and 87,3%,22.7% and 54.1% for soil conservation benefits,respectively.The order of soil and water conservation benefits was millet>soybean>mazie,and the order of benefits on reducing sediment was millet>mazie>soybean.The comprehensive influence of rainfall intense,slope and crop coverage on soil and water conservation could be predicted and simulated through the general formula of ?W=A?PI?^a,S^bC^-c and ?S=A?PI?^a,S^bC^-c.The contribution of crops to soil and water conservation could be described through flow reduction coefficient and sediment reduction coefficient of vegetation which were both affected by rainfall energy,slope and crop coverage,which could be expressed by CW=1.324?PI?^0.33S^0.057C^-0.27 and CW =1.795?PI?^0.105S^0.164C^-0.757.According to calculation,the mean value of water reduction coefficient of these three crops was 0.72,0.73 and 0.76,respectively,and the mean value of sediment reduction coefficient was 0.35,0.45 and 0.35,respectively.3.Rainfall infiltration on the sloping farmland of crops changed in a volatility downward trend as the rainfall duration was extended under artificial simulated rainfall.The infiltration amount was the maximum in the first 10 min of rainfall,and soil infiltration amount was higher than rainfall intensity,so rainfall was infiltrated completely.When surface runoff generated,infiltration capacity tended to a relatively stable stage of infiltration in the fluctuation and reduction process due to the comprehensive influence of rainfall intensity,slope,soil water content,soil crust and crop leaf area index.The mean infiltration capacity of millet,soybean and mazie in the sloping farmland was the 1.1,1.2 and 1.3 times respectively that of the control group,and the initial runoff-generation time was extended by 6.4min,6.18 min and 10.36 min respectively in three crop sloping farmland,which might be the roots cause that autumn crops could prevent soil erosion.The initial infiltration rate,the mean infiltration rate,the stable infiltration rate and the cumulative infiltration amount differed with different crops,and there was significant difference between the internal of the same crop.As for the cumulative infiltration amount,it was 104.6mm,62.64 mm and 36.18 mm respectively in the mazie,soybean and millet sloping farmland,which suggested that crops had a significant effect on the soil water storage.Due to various factors,the maximum,minimum and mean values of four characteristic parameters were obtained with some randomness,but all could be estimated through the corresponding empirical model.Results simulated according to four models by Kostiakov,Horton,Jiang dingsheng and Philip showed that the model by Kostiakov could reflect the soil infiltration characteristics of autumn crop sloping farmland in some extent,which due to the power functions relationship between rainfall infiltration and infiltration time.4.Statistical analysis demonstrated that the stepwise multiple regression analysis and the improved BP neural network method both could describe the characteristics of soil infiltration,and the interaction between main factors and factors was both considered as well.However,when comparing the two,the mean soil infiltration rate obtained by the improved BP neural network method was sufficiently close to the measured value,which could more accurately described the soil erosion under different conditions in study area.The default factor test and the gray degree analysis in BP neural network model both demonstrated that effects of each factors on the soil infiltration were in the order of rainfall>leaf area index>slope.