陕北黄土高原森林群落物种多样性分析

晋陕蒙能源基地榆神府矿区,煤炭资源丰富,是我国重要的能源重化工基地。由于地处鄂尔多斯盆地东北部的黄土丘陵区和毛乌素沙地,自然条件恶劣,生态环境十分脆弱,水土流失是该区主要环境地质问题之一,而煤炭资源的大规模开发,必将加剧水土流失和环境恶化。根据1987年、1992年和1999年3期TM卫星遥感影像解译成果,在地理信息系统空间数据库功能的支持下,应用层次分析法,评价了该区水土流失现状,初步揭示了水土流失影响因素和发展规律。结合调查成果,研究了矿山开发等对水土流失的影响,并进行了初步预测评价,为该区环境保护和水土流失的防治提供了依据。     

不同土地利用模式下土壤侵蚀空间演化模拟

高分子聚合物具有改善土壤结构,增加降雨入渗,调控坡面径流的作用。为探索适合于黄土区应用的高效低廉降雨径流调控新材料,选择聚丙烯酸、聚乙烯醇、脲醛树脂3种高分子聚合物,以不同浓度施入土壤,研究其对土壤物理性质的影响。实验表明：(1)土壤经3种高分子聚合物处理后平均水稳性团粒含量增加17．27％,渗透性能提高41．81％,容重减少11．18％,土壤持水能力较对照提高2．8倍。这说明此3种高分子聚合物均是较好的土壤结构改良剂,尤其是聚丙烯酸效果最为显。(2)根据3种聚合物不同浓度处理后的综合效果,确定了聚丙烯酸、聚乙烯酵、脲醛树脂的适宜浓度分别为4．8％—6．0％,1．6％,10．0％。     

两种土壤水分特征曲线间接推求方法对黄土的适应性评价

土壤水分特征曲线(SWCC)是模拟土壤水分运动和溶质运移的一个重要参数,利用十壤的基本物理性质来间接推求SWCC的方法已经成为当今土壤物理学领域的研究热点.为了比较两种SWCC间接推求方法——Arya-Paris物理经验方法(简称AP方法)和Tyler-Wheatcraft分形几何方法(简称TW方法)对黄土的适应性,该文分析了黄土高原296组土壤颗粒分布、容重和水分特征曲线等资料,利用简化的Fredlund(Fred3P)模型模拟得到连续的十壤颗粒分布曲线,然后应用AP和TW方法预测出相应吸力下的土壤含水量.研究结果表明,对于黄十性土壤,AP和TW两种方法的预测结果均达到了一定的精度,相比较而言AP方法的预测效果明显优于TW方法,且受质地影响小.     

Trends in relief design and construction in opencast mining reclamation

This paper examines trends in topography design and construction in opencast mining restoration. The greatest geomorphological focus is the protection of aquatic ecosystems downstream of reclaimed sites through the construction of topographies and structures that not only reduce sediments and contaminants to a minimum, but which soften the impact of maximum water flows. From the ecological point of view, the most important focus is the integration of the geomorphological component with those of the soil and vegetation, and the formation of a functional ecosystem with the capacity to maintain itself. The conceptual models of relief that have been used in practice are discussed and criticized: these include the platform‐bank model (geotechnically but not ecologically stable), faithful reproduction of the original topography (inadequate in steep areas), and understanding the hydrological basin as a restoration unit (now considered the most appropriate). Finally, erosion models are reviewed as tools for relief design. The practicality of ‘RUSLE 1.06 for Mined Lands, Construction Sites and Reclaimed Lands’ is underlined owing to its ease of use. Copyright © 2003 John Wiley & Sons, Ltd.     

Factors influencing adoption of soil and water conservation measures in southern Mali

A soil and water conservation (SWC) extension programme, promoting erosion control measures and soil fertility measures, has been going on in southern Mali since 1986. Five factors that influence farmer adoption of SWC measures were analysed: land pressure, cotton‐growing area, possession of ploughing equipment, possession of a donkey cart and farmer training in SWC. Interviews were carried out with 298 farmers and two to three fields per farmer were visited, in 30 representative villages and 30 villages with high SWC adoption. Correlation, regression and factor analysis led to the following conclusions:

• (1) Farmers in the high land‐pressure area adopt more soil fertility measures.
• (2) Farmers in the cotton‐growing area adopt less SWC measures.
• (3) Farmers with more ploughing equipment adopt more SWC measures.
• (4) Farmers with a donkey cart adopt more soil fertility measures.
• (5) Trained farmers adopt more erosion control measures.

There is a strong correlation between the adoption of erosion control measures and soil fertility measures that could not be explained by these five factors only. This suggests that there are additional factors that trigger the adoption of SWC measures. Copyright © 2003 John Wiley & Sons, Ltd.     

基于电磁感应数据的电导率反演模型研究

【目的】土壤盐渍化是限制新疆南部棉花高产的主要因子,准确获取区域尺度土壤剖面盐分信息。【方法】以南疆阿拉尔垦区为研究区,以田间尺度采集的30个不同盐渍化程度棉田的540个样点的0~0.375、0~0.750、0~1.000 m的土壤剖面电导率数据和对应的电磁感应数据为数据源,采用线性模型和非线性模型分别构建了田间尺度和区域尺度的土壤剖面电导率的电磁感应反演模型,并采用缩减建模样本量方法进一步检验了区域尺度模型的可靠性和稳定性。【结果】多元线性回归(MLR)、偏最小二乘回归(PLSR)和主成分回归(PCR)建模方法的田间尺度模型R²在0.88~0.95,而对应的区域尺度模型R²在0.34~0.53。基于随机森林(RF)、神经网络(NN)和支持向量机(SVM)非线性建模方法构建的土壤剖面电导率的区域尺度电磁感应反演模型R²在0.60~0.85,其中RF模型的精度最高。0~0.375、0~0.750、0~1.000 m土壤剖面电导率的RF反演模型R²分别为0.80、0.85和0.84,相较于线性建模方法的区域尺度模型精度有明显的提高。RF区域尺度模型的样本数量由540个缩减到240个,模型精度没有明显变化,表明采用区域尺度模型,可大幅度降低土壤剖面样本采集数量,从而可显著提高采样效率和降低采样成本。【结论】区域尺度下构建土壤剖面电导率反演模型时,随机森林建模方法效果较优,模型预测能力具有较高的可靠性。     

物联网模式下的农业土壤养分监测及施肥系统研究

针对目前国内传统施肥方式中,存在基肥施用量大,盲目施肥,氮、磷、钾单一品种化肥用量过多,导致土壤肥力退化、土壤养分不均衡,影响到农作物增收等现状,借助物联网技术、近红外分析技术、现代通信技术和无线网络技术,选用近红外分析技术和基于FFT算法及小波变换的光谱微弱信号处理方法,实现实时监测土壤养分的变化,全面而准确地掌握农业土壤养分情况,并智能控制自动施肥执行端施肥。为此,进行了样机试制,并运用于不同的农业土壤,包括褐土、红土、黑土等进行养分监测和施肥试验。结果表明:农业土壤养分监测及施肥系统可稳定运行,精准施肥,能满足不同农业土壤的精确施肥。本系统可指导科学施肥,提高农业生产的施肥水平,减少农业污染。     

球墨铸铁深松铲尖磨损试验分析

通过对比球墨铸铁与进口铲尖在相同时间、相同工作量下的失重量,探究了两种材质的耐磨性能,验证了球墨铸铁铲尖作为铸钢铲尖替代品的可能性。在相同工作量时,球墨铸铁的耐磨性能与进口铲尖一致。深松技术能提高土壤蓄水保墒的能力,球墨铸铁铲尖作为进口铲尖的替代品,同样能达到深松的效果,使土壤的容重降低并提高土壤的蓄水保墒能力。根据数据显示,深松前土壤含水率为14.29%,深松后土壤含水率为15.30%,深松后土壤平均含水率提高1%,深松前土壤容重为1.92g/cm~3,深松后土壤容重为1.35g/cm~3,深松后土壤容重降低0.57g/cm~3,有利于农作物的生长。