Font Size: a A A

Multi-scale Soil Hydrological Processes And Driving Factors On The Loess Plateau

Posted on:2023-06-07Degree:DoctorType:Dissertation
Country:ChinaCandidate:Y L ZhaoFull Text:PDF
GTID:1523307097998199Subject:Soil hydrology
Abstract/Summary:
Soil water content is a key factor involved in a series of hydrological processes,which determines material transport,the amount of surface and groundwater resources,plant water uptake and usage,vegetation yield,and biogeochemical processes.Soil water content depends on the dynamic changes of input and output and the transformation of various water sources.The key to vegetation restoration and ecosystem service function enhancement of critical zone on the Loess Plateau is efficient and sustainable utilization of limited water resource.Based on the characteristics of spatiotemporal variation and scale dependence of soil water content,and the environmental problems caused by water deficit,the objectives of this dissertation were:(1)to explore the distribution characteristics of soil physical properties at regional and watershed scale;(2)to investigate the spatiotemporal variation of soil water content and dominant controls at regional and watershed scale;(3)to illustrate the effect of intensive land restoration(Gully Land Consolidation)project on soil water storage at watershed scale;(4)to reveal the dynamic evolution process of dried soil layer at watershed scale;(5)to determine the water use strategy of typical vegetation species at plot scale.Based on the research approach of intensive sampling design,field survey,location observation and laboratory analysis,we separately selected 42 and 161 sampling points to collect disturbed soil samples to 5 m depth and undisturbed soil cores in 10 cm and 20 cm soil layer at the regional and watershed scale(treated or not treated by the Gully Land Consolidation project),respectively;moreover,0-5 m soil samples and plant root,plant xylem and rain water were also collected at plot scale.All these samples were used to determine soil particle composition,soil organic carbon,bulk density,saturated hydraulic conductivity,root biomass,hydrogen and oxygen isotopes.We conducted classical statistics,geostatistical,multivariate empirical mode decomposition,empirical orthogonal equation,Mix SIAR model and other techniques to analyze the spatiotemporal characteristics and influencing factors of soil water content,model establishment and prediction of soil water storage,and water use sources of typical vegetation species.The main results are listed as follows:(1)At the regional and watershed scales,saturated hydraulic conductivity showed stronger variation than bulk density.At the regional scale,saturated hydraulic conductivity in grassland was slightly higher than the other land use types.Climate condition had no significant effect on bulk density and saturated hydraulic conductivity.However,land use type exerted significant influence on bulk density variability,while seasons had a significant effect on saturated hydraulic conductivity variability.The sampling time can significantly affect both bulk density and saturated hydraulic conductivity variability.At the watershed scale,Bulk density on slope was significantly smaller than in gully(p<0.05),while saturated hydraulic conductivity on slope was significantly higher than in gully(p<0.05).Bulk density under forest and grassland showed a significantly seasonal variation,whereas saturated hydraulic conductivity displayed a significantly seasonal variation under all land use types.External environment(temperature)was the controlling factor that determined the temporal change of bulk density and saturated hydraulic conductivity.(2)At the regional scale,the analysis of multivariate empirical mode decomposition showed that,along the whole transect,precipitation and temperature were the controlling factors on soil water content at large scales(>250 km),elevation and sand were the factors determining soil water content at moderate scales(65 km),whereas,clay content affected soil water content distribution at all scales.In different climate zones,precipitation and temperature controlled soil water content at large scales in both climatic regions,normalized differential vegetation index had opposite effects on soil water content in the two climate zones,soil organic carbon had a significant impact on soil water content at moderate and large scales in the sub-humid climate zone,clay content had a greater effect on soil water content in the semiarid climate zone compared to the sub-humid climate zone.Scale specific correlations between soil properties and soil water content changed in the five soil layers in the sub-humid climate zone,while similar characteristics were observed among the different soil layers in the semiarid climate zone.Specific scale controlsof associated forcings on soil water content were similar regardless of the soil water condition(dry or wet).(3)At the watershed scale,the result of empirical orthogonal equation showed that the first two EOFs obtained from all locations explained 86%–96% and 87%–97% of total soil water variance in the different soil layers of the treated and untreated watershed,respectively.For all locations,the spatial EOF1 structures matched DEM characteristics in both watersheds.The EOF1 of both watersheds was governed by factors associated with soil water accumulation regimes and soil water storage capacities,while factors that corresponded to solar radiation and energy inputs governed EOF2 for all soil layers of the untreated watershed.Land use type generally played a critical role in determining the dominant spatial soil water structure in both watersheds.Different spatial EOF patterns between all locations and those from slope locations in the treated watershed confirmed the importance of the gully land consolidation project on soil water patterns.The spatial variance percentage for EOF1s(i.e.,PV1s)from all locations showed that they were dominant in all soil layers for both watersheds.Furthermore,different temporal PV dynamics between all locations and those from slope locations in the two paired watersheds indicated the stability of EOF at a watershed scale.(4)At the watershed scale,soil water storage was highly correlated with different local factors in both watersheds,and the minimum dataset contained eight and six factors in the treated and untreated watersheds,respectively.Land use type was the most important factor for predicting soil water storage in the two watersheds.Low soil water storage values occurred on slopes and high soil water storage values were found in the gully in the treated watershed,while soil water storage was pachy distribution in the untreated watershed.The soil water volume declined from spring to summer and increased from summer to autumn in both watersheds.When considering the 5 m soil profile,the ratios of soil water volume in the gully land filled in the treated watershed to the soil water volume of the whole watershed for all four seasons was 22%.Additionally,soil water storage in the two gully-lands displayed similar and weak temporal variation(CV=6% for the new land,CV=11%for the control gully land).Soil water storage were always negatively correlated with mean temperature and wind speed(p < 0.05)with no hysteresis,while soil water storage responded to precipitation in the new land and the control gully land were two and three months hysteresis,respectively.Spatial patterns of soil water storage in the two gully lands displayed distinct difference.The mean proportion of available soil water storage to available water holding capacity was 69% in the new land,indicating the stored soil water was available for crop cultivation without the threat of water shortages.(5)At the watershed scale,dried soil layer(DSL)occurred at approximate 90% and 76% of sampling sites in the treated and untreated watershed,respectively.The spatially and temporally averaged DSL forming depth(DSLFD),DSL thickness(DSLT)and soil desiccation index within DSL(DSL-SDI)in the 0–5 m profile in the treated and untreated watersheds were 170 cm,262 cm,0.49 and 172 cm,283 cm,0.51,respectively.Forestland and shrubland had severer DSL status than grassland in the two watersheds.DSLFD in the two watersheds were sensitive to temperature and precipitation,while DSLT and DSL-SDI exhibited different degree of response to meteorological conditions in the two watersheds.The redundancy analysis(RDA)exhibited that the first two axes accounted for 48.3% and 77% of the total DSL variation in the treated and untreated watershed,respectively.Land use type accounted for 22.3% of the total DSL variation,followed by topographic(7.6%)and soil properties(1.8%),which was determined by the variation partitioning analysis(VPA)in the treated watershed,while land use type,soil and topographic properties explained 18.3%,16.8% and 11.9% of the total DSL variation in the untreated watershed,respectively.Both the RDA and VPA demonstrated land use type was the dominant control of DSL variation in the two watersheds.The comparison of the spatial patterns of DSLT under three conditions in the two watersheds suggested that it is feasible to plant grassland on the upslope with high elevation and mixing plant forestland/shrubland and grassland on middle and down slope with low elevation at watershed scale.(6)At the plot scale,plant roots were mainly distributed within the 0–100 cm soil layer.The soil water content and the stable isotope signatures in the upper soil layers showed significant variation over the transition period.δ18O and δD signatures in precipitation,soil water and plant water exhibited an apparent correlation during the study period.Throughout the transition period(April-July),the trend in water uptake from the shallow soil profile layer(0–40 cm)and the intermediate-shallow soil profile layer(40–120 cm)increased;while the trend in water uptake from the deep soil profile layer(200–500 cm)decreased.This closely correlated to precipitation and plant root distribution.The proportion of water absorption of robinia pseudoacacia varied seasonally for different ages of stand.In the growing season,different ages of robinia pseudoacacia mainly used shallow soil water(0-60 cm),6-year-old robinia pseudoacacia also absorbed much middle soil water(33.1%),while 16-and 20-year-old robinia pseudoacacia absorbed more water from deep soil(31.8% and 31.6%).Compared with 16-and 20-year-old robinia pseudoacacia,35-and 45-year-old robinia pseudoacacia utilized less water from deep soil.In the non-growing season,different ages of robinia pseudoacacia mainly uses deep soil water.DSL occurred in each stand age of robinia pseudoacacia and the most serious DSL was found in 20-year-old robinia pseudoacacia.The development degree of DSL increased with the increasing age of robinia pseudoacacia.The robinia pseudoacacia still use soil water within the DSL,which would aggravate the evolution of DSL.Soil water content is an important variable in soil science,hydrology,agriculture and other disciplines,and its spatiotemporal variation characteristics and scale dependence deeply affect the understanding of soil water movement processes.Knowledge of spatiotemporal variation characteristics,influencing factors,modelling and prediction of soil water and relation with the water use strategy of vegetation is benefit to scientifically assess and manage regional soil water resource.This information is also helpful to clarify the balance between vegetation water consumption and limited water resources and to determine the vegetation carrying capacity of soil water,which has important theoretical and practical significance for the sustainable development of vegetation restoration and ecological reconstruction on the loessPlateau.
Keywords/Search Tags:Spatiotemporal variation, Soil water content, Regional scale, Watershed scale, Multivariate empirical mode decomposition, Empirical orthogonal equation, Stable hydrogen and oxygen isotope, Vegetation water use strategy
Related items