| Yinchuan Plain,located in the arid and semi-arid areas of Northwest China,is a key area for ecological protection and high-quality development of the Yellow River Basin.Due to the influence of human activities such as diverting water from the Yellow River for irrigation,the problem of soil salinization in Yinchuan Plain is prominent,which seriously hindering the high-quality development of regional economy and society.Soil salinization is closely related to shallow groundwater circulation,while how to formulate scientific and reasonable salinization prevention and control measures depends on a deep understanding of groundwater circulation characteristics and soil salinization driving mechanism.Due to the complex spatial and temporal variability in the formation and development of soil salinization,the traditional means are difficult to achieve a wide range of spatial and temporal multi-dimensional research,resulting in the difficulty of precise measures for salinization prevention and control,which has become an urgent problem to be solved for local agricultural economy and ecological protection.In view of this,based on the theories and technologies of hydrogeology,environmental geology,environmental geochemistry and geographic information system,this paper focuses on the following three aspects by means of isotope analysis,groundwater numerical simulation,remote sensing inversion,machine learning,MGWR regression analysis:(1)characteristics of shallow groundwater circulation and soil salt accumulation;(2)study on groundwater salt migration and soil-water response;(3)research on intelligent inversion and driving mechanism of soil salinization.Finally,the prevention and control measures of soil salinization were put forward.The main research results are as follows:(1)Characteristics of shallow groundwater circulation and soil salt accumulationBased on the regional water chemistry analysis and isotope analysis to reveal the characteristics of shallow groundwater environment in Yinchuan Plain,three water circulation models were constructed,and the results showed that Yinchuan Plain is a neutral-weakly alkaline shallow groundwater body,slightly saline,with Na+,SO42-,and Cl-as the main TDS control ions,distributed in Dawukou-Pingluo and the area to the north.The water chemistry and multiple isotope analysis concluded that the shallow groundwater is light in age and fast in renewal rate.There are more types of water chemistry types,overall HCO3-Na,SO4-Na,HCO3-Ca type water is dominant.The water chemistry components are mainly influenced by groundwater circulation,evaporation concentration and human activities.Based on this,the shallow groundwater circulation pattern in Yinchuan Plain can be divided into"Regional lateral discharge type",Local lateral discharge type"and"Vertical discharge type".Based on the regional soil sample testing,multivariate mathematical and statistical analysis and shallow groundwater circulation patterns,the salt accumulation characteristics of the surface soil in the Yinchuan Plain were studied.The results showed that:the salt content in the soil samples ranged from 0.14 to 49.33g/kg,and 54.33%of the samples showed salt accumulation,with a spatial pattern of low in the south and high in the north.Based on the p H and salt content,the regional soil can be broadly classified into:alkaline,saline,saline alkaline,low saline and alkaline.Numerical and statistical analysis shows that K++Na+,Ca2+,Mg2+,Cl-,SO42-are the main components of soil soluble salts,which are salt-based components after evaporation and concentration in typical arid-semi-arid areas.In addition,combined with the shallow groundwater circulation pattern,the soil salinization in Yinchuan Plain can be classified into"Permeable soil salinization","Bottom aggregation type soil salinization"and"Surface aggregation type soil salinization".(2)Research on groundwater salt migration and water-soil responseThe evapotranspiration model of MOD16 is the basis of the water-salt model.The model calculates that the evapotranspiration is increasing in most areas of the study area,especially in the northern area of the plain area.The average groundwater level in the study area will increase by 0.3m in the next 10 years under the existing water diversion strategy,and locally increase by more than 1m in the northern area.TDS is mainly enriched in the western area of Pingluo,Huinong county along the Yellow River.The response of soil salt accumulation to groundwater level was analyzed by fitting the water-salt model with BP neural network regression equation.It is proved that the groundwater level of Dawukou-Huinong area is mainly affected by drainage capacity and leakage coefficient of canal system.The groundwater level in Helan north is mainly affected by canal system leakage.The groundwater level of Xidatan is mainly affected by drainage ditch.Soil salinity is correlated with Na+,SO42-,Cl-,Ca2+,Mg2+and TDS of groundwater,among which Na+,Mg2+and TDS have a great influence on soil salinization.(3)Study on intelligent inversion and driving mechanism of soil salinizationBased on Google Earth Engine(GEE)remote sensing big data cloud platform,an intelligent inversion model of soil salinization was constructed.The results showed that in 2022,soil salinization was more serious in the north than in the south,and in the east than in the west of Yinchuan Plain.Three typical areas salinization was severe.The salinization area of Dawukou-Huinong area is 37.89km2,accounting for95.5%of the area with moderate degree or above,the salinization area of Xidatan was16.32km2,accounting for 92.0%of the area with moderate degree or above and the salinization area of the north of Helan Town was 14.48km2,which accounting for93.2%of the area with moderate degree or above,Statistical display of each county,Pingluo county had the largest area of severe salinization and salinization,Huinong and Dawukou district had the highest proportion of salinization and salinization.In terms of time scale,soil salinization decreased from 2010 to 2022.Salinization and severe salinization decreased significantly,while non-salinization areas increased significantly.On the spatial scale,severe salinization showed a trend from wide distribution to concentrated distribution in the north,and sporadic distribution in the rest areas.Nine driving factors including groundwater salinity,groundwater depth,aqueduct,drainage ditch,evaporation,soil type,slope,precipitation and land use type were selected based on the geographic detector,and a multi-scale geographically weighted regression model(MGWR)was established.The calculated results showed that:The average regression coefficient of groundwater salinity was 0.539,and the groundwater depth was-0.276.The above two factors were the main driving factors of soil salinization.In addition,precipitation is-0.14,evaporation is 0.121,canal system leakage is 0.102,land use type is 0.101,drainage ditch is 0.079.These five factors also have different degrees of influence on salinization.The effect of soil type and slope is not obvious.Then,the driving function of three typical regions was simulated quantitatively,and the driving mechanism of soil salinization was analyzed correspondingly.In the end of this study,based on the formation mechanism of soil salinization in the study area,corresponding water resources development and utilization and soil salinization prevention measures were proposed for three typical areas. |