| Soil salinization is a global ecological environment issue which threatens the ecological environment and sustainable agricultural development.In recent years,the water diversion decreased with the development of water-saving reconstruction projects,it changes the distribution of groundwater,crop planting and salt.In particular,the response mechanism of spatial-temporal evolution of soil salinization to different external factors,the spatial distribution of regional land types and the variation trend of soil water and salt under water-saving conditions are still unclear.Therefore,clarify the influence of various factors on soil salinization and to comprehensively predict the variation trend and spatial distribution of regional soil water and salt under water-saving irrigation is the foundation of making effective salinization prevention and control measures is also the basis of implementing large-scale water-saving reform.Yongji of Hetao irrigation district,a typical salinization region,was chosen as the study region in this paper.The objectives of this study are as follows:1)developing the model of land use type interpretation and soil salinization inversion were constructed based on remote sensing technology on the basis of field sampling,respectively;2)revealing the evolution rule of soil salinization of different land use types in different water-saving transformation periods and determining the threshold value of critical groundwater index for preventing soil salinization;3)determining the appropriate indexes of channel water utilization coefficient and drainage ditch depth under different water management situations used SALTMOD model;4)revealing the future evolution trend of soil salt spatial pattern and its response to water-saving transformation,and determining the optimal layout of main planting structures according to the spatial evolution trend of soil salinization used CLUE-S model.Main conclusions were as follows:1.A decision-tree model was also established to classify the land use types using the integrated band reflectance,spectral index,and texture feature of different periods based on Landsat time-series image data.Which determined the optimal recognition period and optimal combination scheme,and provided a theoretical basis for the selection of single period image in land use type interpretation.The accuracy of the classification model in the band,spectral index,and texture feature using remote sensing images in August was better than that of other periods,where the misclassification was reduced on the spatial layout of salinized wasteland and residential land.Specifically,the overall accuracy,Kappa coefficient,producer accuracy,and user accuracy were 80.23%,0.74%,80.95%,and 86.26%,respectively.Correspondingly,the best identification period was August in the study area,followed by September.The combined feature variables greatly improved the accuracy of classification,where the average overall accuracy and Kappa coefficient increased by 6.72%and 0.09,respectively,compared with the single feature variable.In the study area,the proportion of agricultural land is the largest and showed an increasing trend,and the largest increase rate was 39.54%from 2005 to 2010.With the improvement of water saving,the increase range decreased and stable.The area of salinized wasteland accounted for 4.26%to 12.65%of the study area,and showed a decreasing trend on the whole,which mainly concentrated in the northern and central areas of the irrigation area.2.Compared with single multispectral model,the accuracy of fusion model is significantly improved,which was established with measured hyperspectral and multispectral data by multiple stepwise regression based on characteristic spectral indices.The R2 of the training set and validation set of the hyper-multispectral inversion model with characteristic spectral indices were 0.651 and 0.635 on average,the RMSE were 2.44g/kg and 2.49g/kg on average,respectively,which R2 were 36.19%and 35.64%higher than training set and validation set of the multispectral inversion model,and the RMSE were 34.28%and 41.72%lower than that.3.With the implementation of water-saving reconstruction projects,soil salinization increased first(2000-2010)and then decreased(2010-2021),and showed a spatial pattern of increasing soil salinization from south to north in the study area.The temporal and spatial evolution of soil salinization was quantitatively analyzed.The area of soil salinization and salt storage increased to the peak in 2010(the middle stage of water-saving transformation)and then decreased.From 2000 to 2021,the average total salt storage capacity of 0-20cm and 20-40cm layer was 1.28 million tons and 1.02 million tons,respectively,and the average total salt storage capacity of different soil layers in autumn decreased by 11.00%and 13.60%compared with that in spring.The unit salt storage capacity of 0-20cm and 20-40cm was 0.824kg/m2 and 0.611kg/m2,respectively.Among them,the unit salt storage capacity of agricultural land and salinized wasteland was 0.653kg/m2 and 0.895kg/m2,respectively.The critical threshold of groundwater index to prevent soil salinization were determined.In spring(April),the critical thresholds of groundwater depth of lightly,moderately,severely salinized soil and salinized soil was 2.663m,2.417m,2.292m and 2.037m,respectively.In autumn(October),they was 2.452m,2.280m,2.064m and 1.872m,respectively.The critical groundwater mineralization thresholds in spring(April)was 1736.146 mg/L,2514.267 mg/L,3288.594 mg/L and 3711.378mg/L,respectively.4.The response of regional water and salt to different water management measures was quantitatively analyzed used SALTMOD model.Under the current irrigation and drainage conditions,the soil salinity of 0～40cm in agricultural land decreased year by year,and decreased by 0.51dS/m(4.90%)in 2035,while wasteland increased by 2.10dS/m(13.86%)in 2035.At the same time,the drainage salinity and groundwater depth increased by 0.19dS/m(8.20%)and 0.24m(10.13%),respectively.Considering the simulation results of soil salt content and drainage salinity,the drainage ditch depth should be more than 1.7 meters,and the soil salt content in agricultural land and wasteland could be reduced by 6.83%and 0.87%in 2035.It is suggested that the utilization coefficient of channel water should be controlled between 0.54 and 0.60.When the water diversion remains unchanged,the soil salt content of agricultural land can be reduced by 4.52%in 2035,and the increase of salt content in wasteland can be controlled between 13.86%and 14.59%.When the irrigation amount remained unchanged,the soil salt content of agricultural land could be reduced by at least 4.52%in 2035,and the increase of salt content in wasteland and drainage salinity could be controlled between 12.08%-13.86%and 8.20%-9.49%respectively.5.The development trend of salinization spatial pattern in the future and its response to water-saving reform were quantitatively analyzed used CLUE-S model.Increasing the planting area of wheat and corn would reduce the salinization degree of agricultural land,especially in 0-20cm soil.When the planting area of wheat and corn increased at different growth rates,the salt content of 0-20cm soil in agricultural land could decrease 2.29%-12.06%and 5.65%-25.54%,respectively.The area of non-salinized soil increased by 12.29%-58.37%and 11.78%-55.80%,and that of lightly salinized soil increased by 3.03%-23.21%and 14.48%-69.46%,respectively.Meanwhile,the area of salinized wasteland was increased by 7.43%-34.14%and 4.98%-24.29%.Increasing the planting area of sunflower aggravated the soil salinization of 0-40cm of agricultural land.When the sunflower area increased at different growth rates,the soil salt content of 0-20cm and 20-40cm of agricultural land increased by 8.32%-22.69%and-1.02%-9.76%,the area of severely salinized soil and salinized soil were increased.While the salinized wasteland area decreased by 8.85%-30.99%.The salt content of 0-20cm and 20-40cm increased by 1.87%and 12.60%respectively,and the area of salinized wasteland increased by 40.92%when increasing water-saving force. |
