The Experimental Study And Numerical Simulation Of Groundwater Recharge Processes And Preferential Solute Transport In Soils

Groundwater recharge is an important part in the hydrological cycle.Understanding of precipitation or irrigation induced recharge is critical for evaluation of groundwater resources and assessment of aquifer vulnerability to contamination.Conventional techniques currently used to estimate recharge greatly simplify the complex infiltration-recharge process by assuming that unsaturated soil is homogenous and infiltration water flow is uniform flow.However,uniform flow models often do not correctly describe infiltration processes in field soils because of the presence of hetergenous soil and preferential flow.There is hence a need to use appropriate models based on understanding water flow and solute transport processes to describe groundwater recharge and contaminant transport.In this study a combination of methods based on lysimeters,tracer tests and numerical modeling were used to analyze groundwater recharge processes.A series of lysimeters containing two different types of soil(fine sand and silt loam)and subject to different groundwater levels were used to evaluate precipitation induced groundwater recharge over an 18-year long time period.Short-term tracer experiments under ponded irrigation conditions were also conducted to study the water flow and solute transport processes in the two soils.The recharge process by preferential flow in the silt loam was emphasized in the study.The breakthrough curves obtained from soil columns containing artificial macropores in the lab were also analyzed the characteristics of preferential transport and evaluated the extent of preferential flow based on the time moment analysis and two-region models by CXTFIT.Following observations and conclusions are made:1.A preliminary qualitative(descriptive)analysis of precipitation induced groundwater recharge rates from the large lysimeters,and of associated water contents and measured breakthrough curves,suggest the presence of contrasting infiltration-recharge processes and transport processes in the disturbed(repacked)fine sand and undisturbed more fine-textured silt loam soils.While rainfall and irrigation water uniformly infiltrated into the fine sand and moved towards the groundwater,preferential flow limited to a very small fraction of pore space dominated infiltration and flow in the silt loam lysimeters,and the associated transport processes.2.A comparison of the recharge hydrographs for lysimeters with the same water table depth showed that the silt loam lysimeters always produced higher,more and sharper peaks with earlier arrival than the fine sand lysimeters.This indicated that water flowed downward faster in the silt loam than in the fine sand,and hence that the infiltration-recharge processes in the silt loam lysimeters most likely were dominated by preferential flow.The recharge hydrographs showed in the dry season low rainfall intensity is not enough to initiate preferential flow,and more than 15mm/d rainfall event would initiate preferential flow in the wet season.The increasing process of preferential flow is very quick,and the decreasing process is very slow.The proportion of preferentially infiltration-recharge inside the silt loam column were determined by segregating the recharge hydrograph.During the rainfall season in 2000,the proportion of preferential flow for 2-m,3-m,5-m and 7-m soil column is individually 91.4%,87%,83.5%and 77.3%While during the whole year the proportion is 81.7%,80%,69.2%,and 64.5%o Because the proportion of preferentially infiltration-recharge inside the silt loam column seemed too much the silt loam data from lysimeters such as recharge rate and recharge coefficient should be carefuuly studied before applying to field.3.The prefential flow inside the silt loam column was effected by the rainfall process characteristics to some degree.Preferential flow always happened in the rainfall season.The effects of different rainfall factors on preferential flow were analyzed by SPSS.The results showed that the main factors influencing the amount of preferential flow is rainfall amount,which is significantly positively related to the amount of preferential flow.The main factors influencing the duration of preferential flow are the rainfall duration and rainfall amount.The duration of preferential flow are significantly positively related to rainfall duration and rainfall amount.The main factor influencing the peak of preferential flow is the maxium intensity rainfall,which is positively related to the peak of preferential flow.The lag time of preferential flow was influenced by the water table depth,which is significantly positively related to the lag time of preferential flow.4.The measured breakthrough curves for the two soil types under ponded infiltration tracer experiments on the lysimeters exhibited contrasting shapes.While more or less symmetrical single-peak breakthrough curves far behind the moisture fronts were observed for the fine sand lysimeters,the silt loam lysimeters showed early and multiple-peak breakthrough curves with long tailings.These results again suggest that solute transport in the fine sand likely can be described using the classical convective-dispersive process,while preferential solute transport may need to be considered for the silt loam.The early arrival of the moisture front at the bottom of silt loam lysimeters during the early stages of the tracer experiments similarly points to the presence of preferential flow and transport.During the later stage of irrigations preferential flow is diminished.One possible explaination for this is that as the soil moisture increased during ponded irrigation,the number of preferential pathways decreased due to swelling of the clay soil. The preferential solute transport in the silt loam column had significant implications in the field. For example irrigation water and nutrient moved rapidly through preferential paths so in the field management process some measurement should be achieved to improve the soil structure to diminish the preferential flow or improve the irrigation scheme when the preferential flow is active in the early stage.But in the later stage of irrigation with soil moisture increased the preferential flow diminished the salt will accumulated in the soil with low permeability and low ability of drainage water.Now drainage measurement should be conducted in the field to prevent salt accumulation in the soil.5.The dye tracer experiment in the field showed the preferential flow paths in the silt loam including root,animal caves and dry fissures in the upper soil.But the preferential flow paths in the silt loam columns on lysimeters are coherent.So the extent of preferential flow in the lysimeter is different from that in the field.But the conclusions from qualitative analysis can help to improve the field management process.6.The uniform flow modeling was found to accurately simulate the recharge process, although it slightly underestimated recharge rates during high intensity rainfalls,likely the result of some preferential or unstable flow during relatively wet periods.This preferential flow could be well described with the uniform flow model by using a modified hydraulic conductivity function close to saturation,as shown for the tracer experiment.7.The dual-porosity model used for the flow processes in the silt loam lysimeters accurately captured precipitation-produced recharge and correctly evaluated the total groundwater recharge rate,but at times slightly underestimated daily recharge rates and overestimated the number of recharge peaks during dry seasons.Modeling of the tracer experiment further confirmed that the flow and transport processes in these lysimeters were being dominated by preferential flow during especially the early stage of irrigation.The dual-porosity model did the relatively poor predictions of the diminished preferential flow during the later stages of the irrigations.8.The breakthrough curves(BTCs)obtained from soil columns containing artificial macropores in the lab exhibited different shapes including some with early breakthrough and increased tailing curves which qualitatively indicates the presence of preferential flow and different extent of prefrntial flow in the different soil columns.By time moment analysis first normalized moment(μ₁^′)which yields the mean breakthrough time,the second central moment (μ₂)which quantifies the variance,a measure of the average spread of the BTC,and the skewness(S)which can expresses the degree of asymmetry of BTCs were used as reference indices to denote the extent of preferential flow.The results of time moment analysis showed among all the BTCsμ₁^' of the BTC for soil column containing 2 coherent macropores is lowest,then lower for soil column containing 1 coherent macropore,and high for homogeneous soil column.μ₂ and S had the contrast conclusion toμ₁^'.The results from time moment analysis agreed with conclusions drawn from qualitatively analysis.9.Transport parameters of Cl^- BTCs fitted to a non-equilibrium two-region model were determined using the program CXTFIT.Using these parameters and the time moments of BTCs from soil columns containing artificial macropores and homogeneous soil column,the contribution of preferential flow other than hydrodynamic dispersion and two-region effect to the overall variance of soil columns containing artificial macropores.An integrated parameter,PFSP, defined as the ratio of variance due to preferential flow vis-à-vis variance casued by hydrodynamic dispersion and two-region effect,has been proposed to ascertain or compare the extent of preferential flow.The caculated result showed the value of PFSP is highest for the soil column containing 2 coherent macropores,then higher for the soil column containing 1 coherent macropore,and low for homogeneous soil column.This also indicates the value of PFSP can be used to quantify the extent of preferential flow in the field when the breakthrough curves could be obtained.In conclusion a combination of methods based on lysimeters,tracer experiments and numerical modeling were used to analyze groundwater recharge processes and solute transport processes.The study emphasized on descriptive analysis and modeling the preferential flow and transport in the silt loam column.The conclusions in the study had instructive implications in evaluation of groundwater resources,prediction of groundwater contamination and the field management processes such as irrigation and fertilizer management.