| The Loess Hilly Region of Northwest Shanxi is the key area of soil erosion control and ecological environment construction in China,where the climate is mainly arid and semi-arid,and the water resources are extremely scarce.In recent years,a large number of soil and water conservation and returning farmland to forest(grass)projects have been carried out in this area,which has a strong impact on the local land use pattern and ecological environment in the Loess Plateau.The main manifestation is that the soil of artificial forestland and grassland is generally dry,which not only causes vegetation to decline or even die,but also rapidly deteriorates the soil water-heat environment,which seriously restricts the sustainable development of local agriculture and forestry economy and social people's livelihood.At present,how to scientifically study the soil water-heat characteristics of typical farmland,forestland and grassland in Loess Hilly Region of Northwest Shanxi,reveal the water-heat transfer process and its influencing factors in soil-plant-atmosphere system in this area,and explore the interaction between land use pattern and soil water-heat dynamics has become a serious problem in ecological restoration of Loess Plateau.In this study,three typical land use patterns of farmland(maize),forestland(caragana)and grassland(alfalfa)in Wuzhai County,Shanxi Province were selected as the research objects.Based on field automatic monitoring and laboratory test data,the temporal-spatial variation and equilibrium characteristics of soil water-heat in farmland,forestland and grassland were quantitatively analyzed,and the difference of soil water and heat environment and its influencing factors were revealed.Three databases of soil properties,vegetation growth characteristics and meteorological conditions were established,and the CoupModel was used to simulate the water-heat transfer process of farmland,forestland and grassland in soil-plant-atmosphere system.The local applicability of the CoupModel was discussed in combination with the actual observation data,which provided a technical reference for more accurate judgment of the dynamic variation law of soil water-heat process in Loess Hilly Region.The main findings of this study are as follows:(1)The soil moisture of farmland,forestland and grassland in Loess Hilly Region was divided into five stages according to time change:stable period in winter(Jan.-Feb.),increase period in spring(Mar.-May),shortage period in summer(Jun.-Aug.),increase period in autumn(Sep.-Nov.)and loss period in early winter(Dec.).However,the average soil moisture content in 2016-2017 was grassland(15.12%)>forestland(12.70%)>farmland(11.33%),and it was found that grassland moisture was the highest and relatively stable in the early stage of vegetation restoration,followed by forestland,and this difference was more prominent in months of abundant precipitation.The soil moisture of farmland,forestland and grassland also had different vertical distribution.Soil moisture of farmland decreased with the increase of depth,and soil moisture of forestland and grassland varied with the increase of depth,but all three levels of shallow variation were greater than those of deep.Within 100 cm,the soil layer of farmland from top to bottom had fast-changing layer and active layer,and forestland and grassland also had secondary active layer in addition.The soil of of farmland,forestland and grassland all showed slight,moderate,severe and extremely severe drying levels,but the distribution range of drying in time and space was different.The drying of farmland and forestland was more serious in the shallow soil,while the deep soil drought of grassland was more serious.And the drying of forestland and grassland land was increasing with the number of years,which showed that the drying degree of the same depth soil layer was more serious than before,the drying duration wass longer,and the range of dry layer in the same period gradually extended to the deep.The average moisture storage in the soil layer within 100 cm in 2016-2017 was grassland(151.04 mm)>forestland(126.84 mm)>farmland(113.20 mm).The changes were mainly affected by rainfall and evapotranspiration,so the average moisture consumption(ET)in 2016-2017 was followed by forestland(540.0 mm/a)>grassland(536.0 mm/a)>farmland(522.9 mm).Compared with the average precipitation(P)527.4 mm in the same period,the farmland's ET/P was 99.15%and a slightly 0.85%surplus of precipitation was stored in the soil.Both forestland and grassland consumed more water than precipitation,ET/P was 102.39%and 101.63%respectively,and the water expenditure was higher than income which showed negative balance phenomenon.Long-term excessive water consumption was likely to cause soil drought,and forestland drought risk was the greatest,followed by grassland.(2)The temporal variation about soil temperature of farmland,forestland and grassland in Loess Hilly Region was synchronous,which showed a single peak curve and was divided into two stages:heating(Feb.-Jul.)and cooling(Aug.-Jan.)stage.The temporal variation of soil temperature was the same as that of air temperature,but the soil variation range was relatively moderate,and the lag of individual years was about one month.The average soil temperature for 2016-2017 was followed by forestland(9.5?)>grassland(8.9?)>farmland(8.5?),and the average annual temperature range was grassland(30.0?)>forestland(29.3?)>farmland(28.6?),however there was no significant difference in soil temperature of farmland,forestland and grassland.Although the annual temperature range of farmland was the smallest,the trend of soil temperature change in forestland was the most moderate,followed by grassland.Because of the low thermal conductivity and slow heat transfer,the soil temperature of farmland,forestland and grassland showed obvious vertical gradient with depth:decreased gradually with depth in Mar.-Aug.,remained constant in vertical profile in Sep.,and increased gradually with depth in Oct.-Feb.;and the degree of temperature variation in different soil layers gradually decreased with the increase of depth.The soil temperature of each layer of forestland was the highest,and the average soil temperature of each layer of grassland was higher than that of farmland,but the surface soil temperature of grassland in winter was very low,which affected the water absorption of root system,so the grassland may appear dehydration or water shortage phenomenon in part.There was no significant difference in freeze-thaw process in farmland,forestland and grassland.The freezing process began at the end of November and completely melted at the end of March.The soil freezing process was slow,but the soil ablation process was faster.The soil moisture has strong coupling relationship with temperature,which was negative in the freezing process,and was positive in the ablation process;and the influence of temperature on soil temperature decreased with depth.The soil freezing process was one-way freezing along the soil profile from top to bottom,while the soil ablation process was bidirectional;and the maximum freezing depth of farmland,forestland and grassland may be near or below 100 cm.(3)The simulation results of soil water-heat transfer process using CoupModel model of farmland,forestland and grassland in Loess Hilly Region showed that the simulation effect of soil heat process was better than that of soil moisture.With regard to soil moisture simulation,the CoupModel model had weak ability for soil moisture simulation.The coefficient of determination(R2)of linear regression of soil moisture simulation in farmland,forestland and grassland was 0.01-0.26,the average error(ME)was-3.60-2.67%,the root mean square error(RMSE)was 1.62-6.8%.While the accuracy was not high,the CoupModel model can still show the seasonal fluctuation of soil moisture.In the aspect of soil temperature simulation,the soil temperature simulation values of farmland,forestland and grassland fitted well with the measured values.The R2 was 0.93-0.97,the ME was-2.90--0.36?,and the RMSE was 1.96-3.25?,but the output of soil temperature was relatively small compared with the actual observation,and the simulation ability for deep soil temperature was not as good as shallow.Generally speaking,the CoupModel model had a good applicability in soil temperature simulation in Loess Hilly Region of Northwest Shanxi,and there was a great instability in soil moisture simulation,and the accuracy needed to be improved.According to all the results of this study,the soil heat in Loess Hilly Region of Northwest Shanxi meets the needs of local vegetation growth,and its influence is in the secondary position,and soil moisture is the most important factor that directly affects vegetation restoration and reconstruction Although the soil moisture condition of forestland and grassland in this area is better than that of farmland in the early stage of vegetation restoration and reconstruction,the annual water consumption of forest grassland is greater than that of precipitation recharge,and the long-term excessive development of artificial forest grass will consume a lot of soil water.Therefore,in the vegetation construction,we should follow the way of natural restoration as far as possible,plant baesd on moisture condition,select suitable native tree species and grass species with low water consumption,and take the measures of natural restoration as main and artificial planting as auxiliary.The conclusion of this study can provide important basis for land use planning and plant type selection in vegetation restoration and reconstruction in this area. |
PDF Full Text Request