Greywater reuse for irrigation: Effect on soil properties

A controlled study of the effect of greywater (GW) irrigation on soil properties was conducted. Containers of sand, loam and loess soils were planted with lettuce, and irrigated with fresh water, raw artificial GW or treated artificial GW. Greywater was treated using a recirculating vertical-flow constructed wetland. Soil samples were collected every 10 days for the 40-day duration of the study, and plant growth was measured. Soils were analysed for physicochemical and biological parameters to determine changes caused by the different treatments. It was demonstrated that raw artificial GW significantly increased the development of hydrophobicity in the sand and loam soils, as determined by water droplet penetration time. No significant changes were observed for the loess soil under all treatments. Observed hydrophobicity was correlated with increased oil and grease and surfactant concentrations in the soil. Zeta (ζ) potential of the soils was measured to determine changes in the soil particle surface properties as a result of GW irrigation. A significant change in ζ-potential (less negative) was observed in the raw artificial GW-irrigated sand, whereas no difference was observed in the loam or loess. Soils irrigated with fresh water or treated GW exhibited no increase in hydrophobicity. Fecal coliform bacteria were absent or < 10 CFU g^− 1 in soils irrigated with fresh water or treated GW, but at least 1 order of magnitude higher in raw artificial GW irrigated soils. Only in the last sampling event and only for the loess soil was plant growth significantly higher for fresh water irrigated vs. raw or treated GW irrigated soils. This study demonstrates that treated GW can be effectively irrigated without detrimental effects on soil or plant growth; however, raw GW may significantly change soil properties that can impact the movement of water in soil and the transport of contaminants in the vadose zone.     

Leaching of human pathogens in repacked soil lysimeters and contamination of potato tubers under subsurface drip irrigation in Denmark

The risk for contamination of potatoes and groundwater through subsurface drip irrigation with low quality water was explored in 30 large-scale lysimeters containing repacked coarse sand and sandy loam soils. The human pathogens, Salmonella Senftenberg, Campylobacter jejuni and Escherichia coli O157:H7, and the virus indicator Salmonella Typhimurium bacteriophage 28B, were added weekly through irrigation tubes for one month with low irrigation rates (8 mm per week). In the following six months lysimeters were irrigated with groundwater free of pathogens. Two weeks after irrigation was started, phage 28B was detected in low concentrations (2 pfu ml⁻¹) in leachate from both sandy loam soil and coarse sand lysimeters. After 27 days, phage 28B continued to be present in similar concentrations in leachate from lysimeters containing coarse sand, while no phage were found in lysimeters with sandy loam soil. The added bacterial pathogens were not found in any leachate samples during the entire study period of 212 days. Under the study conditions with repacked soil, limited macropores and low water velocity, bacterial pathogens seemed to be retained in the soil matrix and died-off before leaching to groundwater. However, viruses may leach to groundwater and represent a health risk as for some viruses only few virus particles are needed to cause human disease. The bacterial pathogens and the phage 28B were found on the potato samples harvested just after the application of microbial tracers was terminated. The findings of bacterial pathogens and phage 28 on all potato samples suggest that the main risk associated with subsurface drip irrigation with low quality water is faecal contamination of root crops, in particular those consumed raw.     

On the critical particle size of soil with clogging potential in shield tunneling

Shield tunneling is easily obstructed by clogging in clayey strata with small soil particles. However, soil clogging rarely occurs in strata with coarse-grained soils. Theoretically, a critical particle size of soils should exist, below which there is a high risk of soil clogging in shield tunneling. To determine the critical particle size, a series of laboratory tests was carried out with a large-scale rotary shear apparatus to measure the tangential adhesion strength of soils with different particle sizes and water contents. It was found that the tangential adhesion strength at the soil–steel interface gradually increased linearly with applied normal pressure. When the particle size of the soil specimen was less than 0.15 mm, the interfacial adhesion force first increased and then decreased as the water content gradually increased; otherwise, the soil specimens did not manifest any interfacial adhesion force. The amount of soil mass adhering to the steel disc was positively correlated with the interfacial adhesion force, thus the interfacial adhesion force was adopted to characterize the soil clogging risk in shield tunneling. The critical particle size of soils causing clogging was determined to be 0.15 mm. Finally, the generation mechanism of interfacial adhesion force was explored for soils with different particle sizes to explain the critical particle size of soil with clogging risk in shield tunneling.     

Experimental study on the clogging effect of dredged fill surrounding the PVD under vacuum preloading

Clogging effect surrounding prefabricated vertical drains (PVDs) is a typical problem when vacuum preloading is applied to a dredged fill foundation. A large-scale model test was designed to clarify the cause and mechanism of the clogging effect, and the basic physical and mechanical parameters of the soil in the clogging zone were tracked during the test. The results demonstrated that a clogging zone was formed around the PVD in the early stage of improvement with conventional vacuum preloading, and the boundary of the clogging zone was approximately 0.2–0.4 of the boundary radius. The clogging zone surrounding the PVD was formed because of the overall movement of the soil toward the PVD under the high vacuum pressure gradient, rather than fine particle migration. The soil in the clogging zone exhibited permeability anisotropy and equivalent ‘smear’ effect. The permeability ratio (k_h/k_v) was less than 1, and the ratio of horizontal permeability coefficients at the test distances of 45 cm and 10 cm were 9.6 at a depth of 20 cm and 8.9 at a depth of 80 cm. An analysis of the microstructure of the soil in the clogging zone demonstrated that the clay particles tended to be vertically oriented. The re-orientation of the clay particles reduced the horizontal permeability coefficient and led to the permeability anisotropy of the soil in the clogging zone. Thus, decrease in the horizontal permeability coefficient and equivalent ‘smear’ effect of the soil in the clogging zone affect the consolidation of dredged fill, which leads to the clogging effect. The permeability anisotropy also slightly affects consolidation.     

Algae removal by fine sand/silt filtration

A laboratory scale study was undertaken to determine the potential of a method of filtering algae from water using fine sand/silt as the filter media. Five median sand sizes (0.064–0.335 mm) and four bed depths (3.175–12.700 mm) were examined in constant head experiments with the algae Scenedesmus quadricauda. A total of 46 experiments were conducted with continuous measurements of filtration rate, head loss and effluent quality. All media with median sand sizes at or below 0.200 mm gave consistently high algae removal rates. The average removal was 97.27% (based on fluorescence) with a low average initial head loss across the filter media of 7.3 cm (median grain size diameter of 0.200 mm with the bed depth of 3.175 mm). Initial filtration rates obtained in the experimental apparatus were as high as 226 m³ m⁻²–day⁻¹ (3.84 gpm ft⁻²), comparable to conventional sand units. Run times were short due to surface clogging of the media. No chemical addition was required to obtain high removal levels.     

Hydraulic and purification behaviors and their interactions during wastewater treatment in soil infiltration systems

Four three-dimensional lysimeters were established in a pilot laboratory with the same medium sand and either an aggregate-laden (AL) or aggregate-free (AF) infiltration surface and a 60- or 90-cm soil vadose zone depth to ground water. During 48 weeks of operation, each lysimeter was dosed 4 times daily with septic tank effluent (STE) at 5 cm/d (AL) or 8.4 cm/d (AF). Weekly monitoring was done to characterize the STE, percolate flow and composition, and water content distributions within the lysimeters. Bromide tracer tests were completed at weeks 0, 8, and 45 and during the latter two times, ice nucleating active (INA) bacteria and MS-2 and PRD-1 bacteriophages were used as bacterial and viral surrogates. After 48 weeks, soil cores were collected and analyzed for chemical and microbial properties. The observations made during this study revealed a dynamic, interactive behavior for hydraulic and purification processes that were similar for all four lysimeters. Media utilization and bromide retention times increased during the first two months of operation with the median bromide breakthrough exceeding one day at start-up and increasing to two days or more. Purification processes were gradually established over four months or longer, after which there were high removal efficiencies (>90%) for organic constituents, microorganisms, and virus, but only limited removal of nutrients. Soil core analyses revealed high biogeochemical activity within the infiltrative zone from 0 to 15 cm depth. All four lysimeters exhibited comparable behavior and there were no significant differences in performance attributable to infiltrative surface character or soil depth. It is speculated that the comparable performance is due to a similar and sufficient degree of soil clogging genesis coupled with bioprocesses that effectively purified the wastewater effluent given the adequate retention times and high volumetric utilization's of the sand media.     

Virus removal within a soil infiltration zone as affected by effluent composition, application rate, and soil type

The column studies presented in this paper simulated the infiltrative surface of onsite wastewater systems where effluent is applied and where a biomat may form. Two bacteriophages, MS-2 and PRD-1, were used as surrogates for human pathogenic enteric viruses during two tracer tests. A vacuum manifold was used to simulate the drainage effects of an underlying unsaturated soil profile, allowing for the collection of percolate samples at 4 cm immediately below the infiltrative surface. The impact of effluent applied (septic tank effluent (STE) or a simulated ground water), soil type (medium sand or sandy loam), hydraulic loading rate (5 or 25 cm/day) and method of application (four equivalent daily doses or 24 equivalent micro-doses per day) on the removal of viruses were investigated. These unsaturated mini column experiments demonstrated that the removal of viruses within an infiltrative surface zone (of approximately 4 cm) generally improved over time under the conditions studied. An exception occurred in sand-filled columns dosed with STE where the removal of PRD-1 decreased after a period of effluent application. Statistical analysis conducted on the calculated percent removal demonstrated that the quality of the effluent applied to the infiltrative surface is important for removal of MS-2 and PRD-1. Hydraulic loading rate also proved important in the removal of viruses. At the time of tracer test 2, columns dosed at the higher HLR (25 cm/day) had higher percent removals for both MS-2 and PRD-1. Soil type altered the removal of PRD-1 at the time of the second tracer test, at which time sandy loam had higher removal rates for PRD-1. No significant differences were observed between columns dosed four times daily and those dosed 24 times daily for either bacteriophage at either of the tracer test time points. These data suggest that over a relatively short period of operation the infiltrative surface of soil based wastewater treatment systems can achieve much higher removal then initially measured shortly after startup.     

Methane eruptions from landfill final cover soil during rainfall events in laboratory experiments

Landfill sites are significant sources of greenhouse gases, particularly methane. It is believed that methane gas eruptions occur under a surface ponding condition during rainfall events. In this study, based on laboratory column experiments in which mixed gas was applied to variably compacted soil columns (10 cm in diameter and 30.5 cm in height) under different rainfall intensities, the effects of the bulk density of the landfill cover soil, the gas injection rates, and the rainfall intensities on the gas eruption phenomenon were investigated. The gas pressure and the water content were monitored during rainfall events. The main results were that (1) an episodic gas compression and eruption cycle was observed under a surface ponding condition, (2) gas eruptions occurred at a higher gas pressure in the soil with a higher dry bulk density, and (3) a higher gas injection rate resulted in continuous gas eruptions. The monitoring of the gas eruption flux and gas concentration in the soil columns, using oxygen-mixed gas as an injected gas, showed that longer gas compression led to higher gas fluxes during the gas eruptions. The concentration of gas in the soil increased over time because of the upward gas flow in accordance with the eruptions. The calculation of the oxygen mass balance indicated that the eruptions had slightly promoted the emission of gas from that which had been stored in the soil prior to the rainfall events.     

Effect of clogging of stone column on drainage capacity during soil liquefaction

Stone columns act as vertical drains, and due to their high permeability, allow for the quick dissipation of earthquake induced excess pore water pressure. When water flows through a loose sandy soil, it washes away fine soil particles. The fine sand particles get detached when the hydrodynamic force applied on the soil particles breaks the inter particle bonds between soil grains. These detached soil particles are then migrated by the seepage water. Based on the concentration of the soil particles in the seepage water, these may be captured at the pore constriction of gravels during the flow of water through the stone column. Thus, the clogging of stone column initiates which reduces of the permeability of column. The rate of dissipation of pore water pressure during earthquake is affected due to the clogging of column. In this paper, a mathematical model is proposed to determine the rate of dissipation of pore water pressure of stone column-reinforced ground by considering the clogging effect of column. The result obtained from the proposed model is verified with the available in-situ experimental data. A parametric study is also performed to investigate the effect of different parameters of the proposed model on the clogging of stone column. It is observed that when the permeability ratio, compressibility ratio and area ratio decrease, the possibility of clogging increases. The peak value of the excess pore water pressure ratio can increase up to around 50% due to clogging.     

Numerical parametric investigation of infiltration in one-dimensional sand–geotextile columns

Geotextiles are routinely used in separation and filtration applications. Design of these systems is currently based on saturated properties of the geotextiles and the surrounding soils. However, in the field, soil and geotextile can be in an unsaturated state for much of their design life during which they are essentially hydraulically non-conductive. Periodic wetting and drying cycles can result in rapid and large changes in hydraulic performance of soil–geotextile systems. The writers have reported the results from physical water infiltration tests on sand columns with and without a geotextile inclusion. The geotextile inclusions were installed in new and modified states to simulate the influence of clogging due to fines and to broaden the range of hydraulic properties of the geotextiles in the physical tests. This paper reports the results of numerical simulations that were undertaken to reproduce the physical tests and strategies adopted to adjust soil and geotextile properties from independent laboratory tests to improve the agreement between numerical and physical test results. For example the paper shows that the hydraulic conductivity function of the geotextile must be reduced by up to two orders of magnitude to give acceptable agreement. The lower hydraulic conductivity is believed to be due to soil intrusion that is not captured in conventional laboratory permeability tests. The calibrated numerical model is used to investigate the influence of geotextile and soil hydraulic conductivity and thickness as well as height of ponded water at the surface on wetting front advance below the geotextile and potential ponding of water above the geotextile due to a capillary break mechanism. A simple analytical model is also developed that predicts the maximum ponding height of water above the geotextile based on two-layer saturated media and 1-D steady state flow assumptions. The analytical model is used to generate a design chart to select geotextiles to minimize potential ponding of water above the geotextile. Ponding can lead to lateral flow of water along the geotextile in reinforced wall, slope, embankment and road base applications.     

Fate of the antiretroviral drug tenofovir in agricultural soil

Tenofovir (9-(R)-(2-phosphonylmethoxypropyl)-adenine) is an antiretroviral drug widely used for the treatment of human immunodeficiency virus (HIV-1) and Hepatitis B virus (HBV) infections. Tenofovir is extensively and rapidly excreted unchanged in the urine. In the expectation that tenofovir could potentially reach agricultural lands through the application of municipal biosolids or wastewater, and in the absence of any environmental fate data, we evaluated its persistence in selected agricultural soils. Less than 10% of [adenine-8-¹⁴C]-tenofovir added to soils varying widely in texture (sand, loam, clay loam) was mineralized in a 2-month incubation under laboratory conditions. Tenofovir was less readily extractable from clay soils than from a loam or a sandy loam soil. Radioactive residues of tenofovir were removed from the soil extractable fraction with DT₅₀s ranging from 24 ± 2 to 67 + 22 days (first order kinetic model) or 44 + 9 to 127 + 55 days (zero order model). No extractable transformation products were detectable by HPLC. Tenofovir mineralization in the loam soil increased with temperature (range 4 °C to 30 °C), and did not occur in autoclaved soil, suggesting a microbial basis. Mineralization rates increased with soil moisture content, ranging from air-dried to saturated. In summary, tenofovir was relatively persistent in soils, there were no extractable transformation products detected, and the response of [adenine-8-¹⁴C]-tenofovir mineralization to soil temperature and heat sterilization indicated that the molecule was biodegraded by aerobic microorganisms. Sorption isotherms with dewatered biosolids suggested that tenofovir residues could potentially partition into the particulate fraction during sewage treatment.     

Elimination of COD, microorganisms and pharmaceuticals from sewage by trickling through sandy soil below leaking sewers

To simulate the filtration and/or degradation of trickling sewage from leaky sewers through the non-water-saturated underground, sewage was trickled through sand of 0.4-2mm from the Rhine valley in glass columns of 125 cm length. For the same sewage the chemical oxygen demand (COD) removal was almost independent of low or high trickling rates. The COD removal efficiency varied, however, from 67% to 79%, for sewage from rain and dry weather periods, respectively. The water content of the moist sand increased from initially 80 ml kg(-1) with increasing sewage trickling rates to 108 ml kg(-1) sand. It remained at 108 ml kg(-1) at higher trickling rates higher than 600 ml d(-1). Analyses of effluent of five consecutive 25-cm soil columns revealed that about 50% of the initial COD were filtrated off on top of the sand or degraded in the uppermost 25 cm at varying trickling rates. Another 6-12% of the COD were removed in the following 25-50 cm of sand, whereas almost no further COD removal was seen in the subsequent two or three 25-cm columns. The COD elimination during trickling of sewage through the segmented column (interrupted random flow) was slightly better than that in the non-segmented column. Total and faecal coliform bacteria decreased faster with increasing trickling depth than that of total aerobic or anaerobic bacteria. After a filter/degradation stretch of 125 cm elimination of all bacteria reached 96.2-99.9%. The sewage contained low concentrations of at least 10 different pharmaceuticals or X-ray media. During trickling of sewage through sand, elimination of these compounds by adsorption onto sand and/or biodegradation varied from a complete removal, e.g. Ibuprofen or Naproxen, to almost no removal for several X-ray contrast media. Some of the medicals were removed as effectively as during conventional sewage treatment.     

Factors affecting the fate of dimethoate in soils

For a given application of dimethoate, the loss by leaching increased with soil type in the following order: clay < clay loam < loam < sandy clay loam (S.C.L.) < sand.

In general, the loss through leaching for a single soil type increased with the amount of dimethoate applied. The degree of retention was greatest in the loam and least in the sand. Retention was thought to be affected by physical forces and hydrogen bonding. The loss of the pesticide through evaporation, degradation and irreversible adsorption increased with the dosage applied. It was in descending order of; clay > clay loam > sand > S.C.L. > loam. However, the loss due to evaporation alone was in the order; sand > S.C.L. > clay loam > loam. Biological degradation of dimethoate was of little importance.     

水源热泵回灌困难颗粒阻塞试验研究

 以水源热泵回灌过程中物理堵塞现象作为研究对象,利用玻璃珠和砾石颗粒作为多孔介质,以碱性氧化铝作为悬浮颗粒,利用自行研制的砂层阻塞试验系统装置模拟多孔介质中悬浮颗粒的迁移和沉积过程,试验中介质运动速度控制在20～200 cm/min。研究悬浮颗粒在玻璃珠和砾石颗粒中沉积时水压力与流速关系,对悬浮颗粒在2种不同介质中的沉积机制进行分析;同时研究玻璃珠介质在相同的流速和不同的悬浮颗粒浓度下相对渗透率与时间的变化关系;最后给出渗透系数衰减模型,模型对预见渗透系数的降低是有效的。     

Experimental and numerical analysis of large scale pull out tests conducted on clays reinforced with geogrids encapsulated with coarse material

The pullout test is one of the methods commonly used to study pullout behavior of reinforcements. In the current research, large pullout tests (i.e. 100 × 60 × 60 cm) have been conducted to investigate the possibility of pullout resistance enhancement of clays reinforced with HDPE geogrid embedded in thin layers of sand. Pullout tests on clay–geogrid, sand–geogrid and clay–sand–geogrid samples have been conducted at normal pressures of 25, 50 and 100 kPa. Numerical modeling using finite element method has also been used to assess the adequacy of the box and geogrid sizes to minimize boundary and scale effects. Experimental results show that provision of thin sand layers around the reinforcement substantially enhances pullout resistance of clay soil under monotonic loading conditions and the effectiveness increases with increase in normal pressures. The improvement is more pronounced at higher normal pressures and an optimum sand layer thickness of 8 cm has been determined for maximum enhancement. Results of numerical analysis showed the adequacy of the box and geogrid length adopted as well as a relatively good agreement with experimental results.     

Model test of clogging effects on composite foundation of geosynthetic-encased steel slag column

Experiments were conducted to study the performance of geosynthetic encased steel slag column with a diameter of 150 mm and a length of 900 mm in a soft clay foundation. The effect of clogging was simulated by mixing the slag with 10% and 20% fines. The measured bearing capacity of the column treated foundation is notably increased to about 10 times than that of the untreated foundation, and is seldom affected by the intrusion of fines. The vertical stress within the soil at column tip attenuates to 85% and 60% of the stress close to the ground surface for the column with no fines and 20% fines, respectively. For the cases with fines content of 10% and 20%, the maximum excess pore pressure is in average 5% and 10% greater than the case without fines, respectively, and the dissipation rate of excess pore pressure is in average 18% and 24% slower than the case without fines. The column treated foundation prevents the water ponding on the surface as that occurs for the untreated foundation. The undrained shear strength of the soil close to the column increases by 18% at the depth of 100 mm, and 6% at the depth of half column, regardless of the fines.     

基于土壤物理特性扩展技术的土水特征曲线预测方法

土水特征曲线是模拟水和污染物在非饱和土中运移的重要水力特性参数。但是,土水特征曲线的直接量测方法比较困难。Arya和Paris提出了一种通过粒径分布曲线预测土水特征曲线的模型——AP模型。该模型引入一个转换系数α建立土体假想形态与真实形态之间的联系。但是,现有的推导系数α的方法一方面计算过程过于复杂,另一方面没有全面考虑土的物理特性。基于土壤物理特性扩展技术提出一种新的计算参数α的方法。为了验证新方法,从非饱和土水力特性数据库中选出不同类型的土壤样本,采用新方法分别计算出各类型土壤的参数α。然后,将计算出的参数α用于预测其他土样的土水特征曲线,从而验证新方法计算出的参数α的有效性。还将提出的新方法与其他利用AP模型预测土水特征曲线的代表性方法进行对比,结果显示该方法预测结果更加准确。     

Removal of trace organic chemicals in onsite wastewater soil treatment units: A laboratory experiment

Onsite wastewater treatment is used by 20% of residences in the United States. The ability of these systems, specifically soil treatment units (STUs), to attenuate trace organic chemicals (TOrCs) is not well understood. TOrCs released by STUs pose a potential risk to downstream groundwater and hydraulically-connected surface water that may be used as a drinking water source. A series of bench-scale experiments were conducted using sand columns to represent STUs and to evaluate the efficacy of TOrC attenuation as a function of hydraulic loading rate (1, 4, 8, 12, and 30 cm/day). Each hydraulic loading rate was examined using triplicate experimental columns. Columns were initially seeded with raw wastewater to establish a microbial community, after which they were fed with synthetic wastewater and spiked with 17 TOrCs, in four equal doses per day, to provide a consistent influent water quality. After an initial start-up phase, effluent from all columns consistently demonstrated >90% reductions in dissolved organic carbon and nearly complete (>85%) oxidation of ammonia to nitrate, comparable to the performance of field STUs. The results of this study suggest STUs are capable of attenuating many TOrCs present in domestic wastewater, but attenuation is compound-specific. A subset of TOrCs exhibited an inverse relationship with hydraulic loading rate and attenuation efficiency. Atenolol, cimetidine, and TCPP were more effectively attenuated over time in each experiment, suggesting that the microbial community evolved to a stage where these TOrCs were more effectively biotransformed. Aerobic conditions as compared to anaerobic conditions resulted in more efficient attenuation of acetaminophen and cimetidine.     

Effects of molecular weight of natural organic matter on cadmium mobility in soil environments and its carbon isotope characteristics

We investigated the role of natural organic matter in cadmium mobility in soil environments. We collected the dissolved organic matter from two different types of natural waters: pond surface water, which is oxic, and deep anoxic groundwater. The collected organic matter was fractionated into four groups with molecular weights (unit: Da (Daltons)) of <1 x 10(3), 1-10 x 10(3), 10-100 x 10(3), and >100 x 10(3). The organic matter source was land plants, based on the carbon isotope ratios (delta(13)C/(12)C). The organic matter in surface water originated from presently growing land plants, based on (14)C dating, but the organic matter in deep groundwater originated from land plants that grew approximately 4000 years ago. However, some carbon was supplied by the high-molecular-weight fraction of humic substances in soil or sediments. Cadmium interacted in a system of siliceous sand, fractionated organic matter, and water. The lowest molecular weight fraction of organic matter (<1 x 10(3)) bound more cadmium than did the higher molecular weight fractions. Organic matter in deep groundwater was more strongly bound to cadmium than was organic matter in surface water. The binding behaviours of organic matter with cadmium depended on concentration, age, molecular weight, and degradation conditions of the organic matter in natural waters. Consequently, the dissolved, low-molecular-weight fraction in organic matter strongly influences cadmium migration and mobility in the environment.     

压实风积沙土层盐分迁移规律研究

针对沙漠地区风积沙路基发生盐渍化病害现象,探究压实风积沙土层水、热、盐变化规律。自制下垫面为硫酸钠盐渍土地基的风积沙土柱试验设备,以自然环境条件下6个月(6月—12月)的实时监测数据为基础,分析压实风积沙土层内水、热、盐的时空分布规律与迁移特征。研究结果表明:受外界环境温度影响,压实风积沙土层分影响剧烈区(0～50cm)与薄弱区(50～120cm),随着埋深增加,土柱内部温度变化幅度趋于平缓,各层温度到达峰值具有滞后效应;受含盐下垫面与压实素土风积沙含盐量的浓度梯度作用,压实风积沙土层深部区域被盐分迁入,由无到有。风积沙土层内盐分的存在,为盐分的进一步迁移奠定了物源基础。由外界温度影响,土层浅部受水分赋存形态的变化与蒸发牵引作用,促使水分携带盐分在压实风积沙土层内部逐渐向上发生迁移;盐分在土层整个埋深断面上呈现"高—低—高"的分布特征。受外界环境荷载周而复始作用,压实风积沙土层浅部盐分将渐渐积聚,由少到多,随着时间推移将形成盐渍化病害。