【作者】 王健；

【导师】 吴发启；

【作者基本信息】 西北农林科技大学 ， 水土保持与荒漠化防治， 2008， 博士

【摘要】 黄土高原土壤水分状况是制约该区生态恢复的关键,也是维持农业持续发展的决定性因素。本文以延安市宝塔区的高坡村、淳化县泥河沟流域和米脂县泉家沟流域为观测区,通过对不同土地利用方式下土壤水分的定位监测,采用定量与定性、模型模拟与实测对比相结合的方法,系统分析了陕北黄土高原土壤水库的特征,取得了以下主要结论:黄土高原深厚的土层,使得土壤水库具有巨大的库容,但受到降水入渗补给的影响,经常呈现出有库无水的状态。受土壤含水量变化及测量误差的影响,实测各层土壤含水量呈现波动变化,以层土壤含水量波动变化最小值对应深度作为土壤水库的调节深度,陕北黄土高原阳坡果园和阳坡刺槐林土壤水库的调节深度最小,阴坡草地和农地土壤水库调节深度最大,其余坡向和土地利用类型土壤调节深度介于此二者之间;土壤水库的调节能力以阴坡农地和荒草地最大,为1194.96mm,阳坡果园和刺槐林最小为597.48mm,其他坡向和土地利用类型介于此二者之间。受地带性、土地利用、坡位、坡向、土地利用年限和土地利用强度等多种因素的影响,坡面土壤水库呈现不同的蓄水特征。在地带性方面,由南向北,土壤储水量明显减少;在土地利用方面,农地和荒草地土壤储水量明显高于果园和刺槐林地;在坡向上,阳坡、半阳坡明显低于阴坡和半阴坡;同一坡向,坡上部含水量低于坡中部低于坡下部;土地利用年限增加,储水量降低,刺槐林16龄林地,200cm以下土壤明显干化;同时植被密度也直接影响着土壤储水量,随密度增加,储水量降低。土壤水库储水量受降水、蒸发等因素的影响,在年内和年际间呈现波动变化;应用时间序列理论,果园、刺槐林、草地和农地土壤水库储水量趋势项均值分别为250.16mm、295.02 mm、284.75 mm和355.86 mm;土壤水库季节项变化规律与土壤水库储水量动态变化规律均为:秋冬季土壤水分缓慢累积阶段、春末夏初土壤水分强烈消耗阶段和夏季土壤水分波动阶段;土壤水库调节着区域降水与蒸散耗水,年内为上半年供给植物生长耗水,储水减少,下半年接受降雨入渗补给,蓄存降水;年际间2003年和2005年土壤水库得到降水入渗补充,储水量增加,2002年、2004年和2006年土壤水分消耗,实测果园、刺槐林、草地和农地土壤水库最大调节量分别达到245.38mm、328.18mm、284.78mm和261.62mm。根据植物从土壤中吸取水分的难易程度,将土壤水分划分为无效水、难效水、中效水、易效水和饱和水。坡面果园、刺槐林、农地和荒草地土壤水分的有效性由南向北递减,表现为淳化大于延安大于米脂;在不同坡向上,土壤水分有效性表现为阳坡最差、阴坡最好,半阴坡、半阳坡和峁顶介于二者之间。年内土壤水分的有效性与季节同步,1月土壤水分整体较高,随土壤水分的蒸散消耗,有效性逐渐降低,难效水范围逐步扩大,6月份表层0~40cm呈现无效水,后期受降水补给影响,土壤水分有效性得到一定恢复。以各层土壤中根系所占份额作为权重,求得各种土地利用类型土壤水分生态位,各种土地利用类型土壤水分生态位适宜度在淳化最高,米脂最低,延安介于二者之间;在各坡向间土壤水分生态位适宜度阳坡最低、阴坡最高,半阴坡、半阳坡和峁顶介于二者之间;在不同土地利用中,以农地土壤水分生态位适宜度最高,刺槐林较高,草地和果园相对较低,各种土地利用土壤水分生态位适宜度在7月达到最低,随后降雨入渗补给,其土壤水分生态位有了一定提高;坡面刺槐林林龄越长,土壤水分生态位适宜度越低,林木的生长受到一定的限制;林分密度越大,土壤水分生态位适宜度越低。应用线形扰动理论,构建土壤水库储出水量线形扰动模型,将降水量和参考作物蒸散量的随机项作为输入,模拟土壤储水量的随机项,叠加土壤水库储水量的趋势项和周期项,构成土壤水库储水量对降水和蒸发的响应;根据降雨、参考作物蒸散量与土壤水库储水量间的互相关关系,确定出土壤水库线形扰动模型各输入的记忆长度为12个时段,即6个月。经验证线性扰动模型可较好的描述土壤水库储水量。利用模型对2006年土壤水库储水量进行预测,经分析模型预测精度除个别时段误差稍大外,其余时段模型模拟值与实测值间偏差很小,模型可较好的预测土壤水库储水量变化。更多还原

【Abstract】 Soil moisture condition is the key which restricts the ecological restoration of the area ,and is also the determinate factor which maintains the sustainable development of agriculture in Loess Plateau. Gaopo village in Baota District in Yan’an, Nihegou Watershed in Chun’hua and Quanjiagou Watershed in Mi’zhi are taken as the area of coverage in this paper. Through position monitoring soil moisture in different land utilization, the character of soil reservoir is analyzed systematically in Loess plateau in northern Shaanxi Province by using the methods of fixed quantity and fixed quality and combination of model simulation and observation comparison. The conclusions are obtained as follows:Submerged soils in Loess plateau, which bring a huge storage capacity into the soil reservoir, are usually in anhydrous condition due to the effect of precipitation infiltration and supply. The change of measured soil moisture content of each layer is fluctuant as it’s affected by the soil moisture content change and inaccuracy of measurement. When the depth that corresponds to the minimum of the soil moisture content fluctuant change is taken as the adjusting depth, the adjusting depth of soil reservoir in sunny slope orchard and sunny slope locust in Northern Shaanxi Province is the least, the shady grassland and farmland is the largest, and the soil adjusting depth of the other slope aspects and land utilizations is intervenient. The modulability of soil reservoir in shady slope farmland and wild grassland is the largest, that is 1194.96mm, the sunny slope orchard and locust is the least, that is 597.48mm, and the modulability of the other slope aspect and land utilization is between them.Domatic soil reservoir have different character as it is affected by multifactor such as zonation, land utilization, slope location, slope aspect, the length of land use and land-use density ect. In zonation aspect, soil water storage grows downwards obviously from the south to the north. In land-use aspect, soil water storage in farmland and wild grassland is obviously higher than that in orchard and locust. In slope aspect, sunny and semisunny slope are obviously lower than shady and semishady slope. In the same slope aspect, water content in the upper is lower than the middle and the abdominal. Soil gets dry obviously under the 200cm soil layer in the locust with the age of 16 along with the length of land use being increasing and water storage was being decreasing. At the mean time, the vegetation density also affects the soil water storage directly. Water storage decreases with the dengsity increasing.Soil reservoir storage is affected by the factor such as precipitation, evaporation and so on , which presents fluctuant change annually and interannually. With the time series theory application, soil reservoir capacity tendency mean value in orchard, locust, grassland and farmland are 250.16mm, 95.02mm, 84.75mm and 55.86mm respectively. The discipline of soil reservoir seasonal change is that: soil moisture is in the phase of slow cumulating in autumn and winter, in the phase of strong consumption in late spring and early summer, and in the phase of fluctuant change in summer. Soil reservoir is adjusting precipitation and evapotranspiration, showing that water storage, which provides plants for their growth consumption, decreases in the first part of year, and that it is supplied by infiltration in the second part of year, storage precipitation. Soil reservoir gets supply from the infiltration in 2003 and 2005, which makes the water storage increasing. Soil moisture is consumed in 2002, 2004 and 2006 when the measured soil reservoir maximum regulated quantities in orchard, locust, grassland and farmland are 245.38mm, 328.18mm, 284.78mm and 261.62mm respectively.Soil moisture is divided into unavailable water, difficult available water, middle available water, readily available water, and saturation water, according to the difficulty level of plants drinking in moisture from the soil. The validity of soil moisture in domatic orchard, locust, farmland and wild grassland recedes into the north, that is Chun’hua > Yan’an >Mi’zhi. In different slope aspects, It’s showed that the validity of soil moisture registers in sunny slope is the worst, shady slope the best, and semishady, semisunny slope and hilly top is between them.Annul soil moisture keeps in step with the season, showing that soil moisture is high integrally in Jan. and validity decreases gradually with the evapotranspiration of soil moisture, and that soil moisture in the 0-40 surface layer is unavailable water in Jun. and validity of soil moisture is recovered in a certain extent for the effect of precipitation supplement. Econiche of soil moisture of different land use type is obtained by taking portion of root in each soil layer as weight, which shows that in different land use types, soil moisture niche fitness in Chun’hua is the highest, Mi’zhi the lowest and Yan’an between them, and that in different slope aspects, soil moisture niche fitness in sunny slope is the lowest, shady slope the highest and semishady, semisunny slope and hilly top between them, and that in deferent land utilization, soil moisture niche fitness in farmland is the highest, locust higher, grassland and orchard lower and that soil moisture niche fitness of all land utilizations reaches the lowest in July, after which soil moisture econiche gets a raise due to the infiltration supplement, and that the longer age of domatic locust is, the lower soil moisture niche fitness is, which restricts the growth of wood in a certain extent and that the larger the density of forest is, the lower soil moisture niche fitness is.With the application of linear disturbing theory, soil reservoir water yield linear disturbing model is built, taking the random value of precipitation and referenced crop evapotranspiration as output , simulating the random value of soil water storage, adding the tendency and periods of soil reservoir water storage and forming the response of soil reservoir water storage to precipitation and evaporation, and make sure that each memory length of soil reservoir linear disturbing model is 12 periods of time, namely six months. Empirical linear disturbing model can describe, soil reservoir water storage very well. Soil reservoir water storage in 2006 is predicted by using the model which can predict the change of soil reservoir water storage very well, for the deviation between simulated value and measured value in other periods is very small through analyzing the forecast precision of the model, except for in individual period when the error is largish.更多还原