不同插值方法对成分数据空间预测结果的影响——以土壤连续分类模糊隶属度值为例

地球科学中成分数据(compositional data)非常普通,其在进行空间插值时必须满足4个条件:每一位置各组分之和为常数,每一组分为非负,插值结果无偏最优.本文以土壤连续分类模糊隶属度值为例,数据经对数正态变换、非对称对数比转换、对称对数比转换后进行普通克里格插值结果和成分克里格插值(compositional kriging)结果进行比较.结果表明,对原始数据和经对数正态变换后数据进行插值,每一位置预测结果隶属度之和不能满足常数1.经非对称对数比转换后,插值结果虽然满足各个位置组分之和为1,但是预测结果精度较低,且预测结果空间分布连续性不明显.数据经对称对数比转换后插值结果和成分克里格插值结果,都能满足成分数据空间插值的4个条件,但二者各有优势.相比较而言,对称对数比转换方法得到的预测结果更能体现土壤空间连续渐变特征,而成分克里格插值结果能保证隶属度本身是最优无偏估计.     

冲积平原上区域土壤质地层次的某些分布特征

本文根据冲积平原区上一个１５平方公里区域内格网化布点的大量土壤剖面质地层次的观测数据,采用地统计学方法研究区域粘土层一些指标值的空间变异结构,通过概率统计研究了区域内几种地质类型层次厚度的概率分布特征。研究结果表明,第一粘土层厚度,第一粘土层出现深度,１米内粘土层总厚度和２米内粘土层总厚度的最大相关距离均在１５００米－２０００米之间,各种类型质地层厚度的概率分布均具有对数正态分布的特点。     

烟田土壤质地的空间变异性研究

为了解烟田土壤质地的空间变异情况进而为烟草精准施肥提供依据,本研究在地统计学和GIS的支持下,以半方差函数为基本工具,分析了平顶山地区典型烟区耕层土壤质地的空间变异特征,并运用普通克里格(Kriging)进行最优无偏线性插值,制作了黏粒、砂粒和粉粒的空间分布图。结果表明,研究区域内不同土壤颗粒在较大范围内存在空间自相关性,黏粒、砂粒和粉粒的空间自相关距离分别为609m、657m和429m;黏粒和砂粒由结构性因素引起的空间变异达80%。各向异性分析都表明,黏粒和砂粒存在较强烈的各向异性,粉粒各向异性则较小。对试验土壤来说,土壤质地具有较强的空间自相关性和变异性,这可能是影响土壤养分空间变异性的主要原因,因此在设计施肥方案时应考虑土壤质地空间变异性的影响。     

应用土壤质地预测干旱区葡萄园土壤饱和导水率空间分布

田间表层土壤饱和导水率的空间变异性是影响灌溉水分入渗和土壤水分再分布的主要因素之一,研究土壤饱和导水率的空间变化规律,有助于定量估计土壤水分的空间分布和设计农田的精准灌溉管理制度。为了探究应用其他土壤性质如质地、容重、有机质预测土壤饱和导水率空间分布的可行性,试验在7.6 hm2的葡萄园内,采用均匀网格25 m×25 m与随机取样相结合的方式,测定了表层(0～10 cm)土壤饱和导水率、粘粒、粉粒、砂粒、容重和有机质含量,借助经典统计学和地统计学,分析了表层土壤饱和导水率的空间分布规律、与土壤属性的空间相关性,并对普通克里格法、回归法和回归克里格法预测土壤饱和导水率空间分布的结果进行了对比。结果表明:1)土壤饱和导水率具有较强的变异性,平均值为1.64 cm/d,变异系数为1.17;2)表层土壤饱和导水率60%的空间变化是由随机性或小于取样尺度的空间变异造成;3)土壤饱和导水率与粘粒、粉粒、砂粒和有机质含量具有一定空间相关性,而与土壤容重几乎没有空间相关性;4)在中值区以土壤属性辅助的回归克里格法对土壤饱和导水率的预测精度较好,在低值和高值区其与普通克里格法表现类似。研究结果将为更好地描述土壤饱和导水率空间变异结构及更准确地预测其空间分布提供参考。     

基于回归克里格和遥感的紫色土区土壤有机质含量空间预测

精确预测紫色土区土壤有机质含量的空间分布,对于指导紫色土区农业生产和培肥土壤具有重要意义。以杜家沟小流域为研究区,以遥感影像作辅助变量,采用回归克里格法,预测土壤有机质含量的空间分布,并与参照方法的预测精度进行比较。结果表明:(1)Landsat ETM+影像的波段2和波段5是土壤有机质含量多元线性回归预测的最佳辅助变量,回归残差的最优半方差函数模型为球状模型,模型的拟合精度较高;(2)土壤有机质含量呈由沟谷逐渐向坡顶递减的趋势,空间变异的细节信息表达较好;(3)回归克里格法在验证点的预测值与实测值的拟合能力更好,预测结果更倾向于无偏的,MAE、RMSE和R2均优于参照方法。因此,回归克里格法是紫色土区土壤有机质含量高精度空间预测的有效方法。     

不同粒径制间土壤质地资料的转换问题研究

使用不同土壤类别的黄绵土、灰褐土机械组成资料 ,通过对比三次样条插值、二次样条插值、线性插值法插值结果 ,探讨了不同粒径制间土壤质地的数学转换方法 ,结果表明三次样条插值法是最优方法 ,其最小相对误差不到 0 6% ,最大相对误差不超过 3 5 % ,平均 2 48%。土壤质地转换数学方法的建立对于应用土壤普查资料计算土壤可蚀性K值 ,诺谟图的验证推广有重要意义     

冀西山前冲积扇区土壤机械组成模型制图对比研究

探索适合地形平缓的山前冲积扇地区土壤机械组成的空间预测方法。以河北省灵寿、行唐、曲阳县400 m高程以下区域为研究区,结合地形因子、土壤类型、归一化植被指数、地表温度等环境变量,选择基于对称对数比（SLR）转换的普通克里格法（SLR-OK）、回归克里格法（SLR-RK）、随机森林法（SLR-RF）3种方法,对训练集114个样点表层土壤机械组成的空间分布进行预测,并通过验证集50个样点比较了3种方法的预测精度。（1）从空间预测图来看,砂粒呈现出西北低、东南高的空间分布趋势;粉粒和黏粒与砂粒相反。与SLR-OK法相比,SLRRK法和SLR-RF法能够更好地反映局部变异并减小平滑效应。（2）对于砂粒和粉粒,SLR-RF法对验证集含量预测的平均绝对误差（MAE）和均方根误差（RMSE）均低于其他两种方法,且决定系数最高,表明SLR-RF的预测精度最高;对于黏粒,SLR-OK法对验证集含量预测的MAE和RMSE均低于其他两种方法,且决定系数最高,表明SLR-OK法的预测精度最高。（3）线性回归预测模型的辅助变量包括高程、土壤类型和风力作用指数;随机森林法模型的辅助变量包括高程、土壤类型、归一化植...     

沂蒙山区降雨侵蚀力空间分布推算方法

以沂蒙山区88个雨量站点1980—2010年日降水数据为基础,采用普通克里格法和回归克里格法对多年平均降雨侵蚀力进行空间插值预测,分析评价不同空间插值方法预测结果的精度及差异性,掌握沂蒙山区降雨侵蚀力及空间分布特征。结果表明:1)2种方法降雨侵蚀力插值结果空间分布规律一致,但回归克里格法对局部变异性描述更为精确;回归克里格法降雨侵蚀力预测值与实算值接近,降雨侵蚀力预测相对精度较普通克里格法提高53.64%。2)沂蒙山区降雨侵蚀力最大值为5 438.22 MJ·mm/(hm2·h·a),最小值3 033.23 MJ·mm/(hm2·h·a),均值4 133.92 MJ·mm/(hm2·h·a);半方差函数分析显示,年均汛期降雨量和年均侵蚀性降雨量是影响降雨侵蚀力空间变异的主要因素。3)沂蒙山区降雨侵蚀力空间分布上遵循从南向西北和东北2个方向逐渐递减的特征,降雨侵蚀力在沂蒙山区空间分布上属中等变异,但在县域尺度空间分布上呈弱变异(邹城除外)。     

Vis-NIR光谱信息辅助的土壤质地协同克里格预测制图

土壤质地传统获取方法费时费力,土壤反射光谱信息是土壤综合信息的反应,以光谱信息为辅助变量的土壤质地协同克里格预测制图值得研究。利用采样点的土壤光谱数据主成分分析综合得分作为协同克里格协变量,并结合GS+软件提供的最佳地统计模型预测土壤质地的空间变异。结果表明以光谱信息为辅助变量的协同克里格方法能提高制图精度,减少样本数量。在主变量减少至60个的情况下,砂粒,粘粒含量的预测值和实测值的均方根误差分别为7.85%,3.03%。     

不同阈值下土壤盐分的空间变异特征研究

指示克里格法(Indicator Kriging,IK)应用的关键是阈值的合理选择。本文以黄淮海平原盐渍土改良区典型县域禹城市为研究区域,在1.0 g kg-1、2.0 g kg-1、3.0 g kg-1等3个盐分阈值下,对0～20 cm耕地土壤盐分的变异函数、预测概率、预测概率空间分布的变化规律进行了研究。结果表明,(1)不同阈值下土壤盐分含量均具有中等强度的自相关性,随着阈值的减小,土壤盐分的空间结构性增强,变异函数的模型精度增大,因此从指示变异函数模型精度考虑,1.0 g kg-1盐分含量为研究区域盐渍化风险评价的最佳阈值;(2)土壤盐分的预测概率最大值和预测概率均值随盐分阈值的减小而增大,可为不同土壤盐渍化风险评价目标下的阈值选择提供参考;(3)不同阈值下土壤盐分的概率预测分布存在空间上的规律性与相似性,高概率区域主要集中在研究区域的西部,低概率区域主要集中在研究区域的东部。研究区域土壤盐分含量的概率分布与地形地貌特征和河流的分布状况有着密切的关系。     

Estimation of soil texture from the sample density

Abstract

Soil texture often plays an important role in the interpretation of soil analytical data for fertilizer advisory purposes. A reliable and inexpensive method of clay content estimation is, therefore, a requirement of most advisory laboratories. This note discusses the use of sample density (i.e. the mass of a scooped volume of soil) as an index of clay content. A strong relationship was found to exist between sample density and clay content, and such estimates of clay content were superior to those obtained by experienced pedologists using the “finger test” procedure. The use of this quick and simple procedure is considered to be ideally suited to soil testing laboratories handling large numbers of samples.     

Using soil E horizon in salvaged topsoil material – effect on soil texture

Soil is a limited natural resource that needs to be efficiently salvaged during landscape construction operations for its further use as topsoil. To avoid inclusion of undesirable subsoil material (e.g. excess clay from B_t horizon), the majority of current guidelines define borrowed topsoil material as the surface layer of native soil, or the soil A horizon. Using information from over 7000 soil pedons from the 48 contiguous United States, we characterized selected topsoil properties and simulated the mixing of A and E horizons. The selected soil properties were compared among four different operationally defined topsoils: A surface layer, A_SL; A_P surface layer, A_PL; A horizon, A; and a mix of A and E horizons, AE. Average topsoil depth decreases in the order: AE > A > A_PL > A_SL; sand content decreases in the order: A_SL > AE > A > A_PL; clay decreases in the order: A_PL > A > A_SL > AE; and organic carbon decreases in the order: A_SL > A > A_PL > AE. On average, mixing of A and E horizons increases excavation depth by over 2.5 fold while having minor effects on soil texture; with AE/A ratio of 1.03, 1.00 and 1.07 for sand, silt and clay content, respectively. Yet, average soil organic matter content decreases by 38% upon mixing A and E horizons. Given the marked increase in soil volume and minor effect on soil texture, it is our suggestion that, for landscaping purposes, protocols for salvaging excavated soil material, for reuse as topsoil material include the E horizon (where it exists). Supplementing the recovered soil material with organic matter, such as compost, to overcome its dilution due to the incorporation of E horizon is recommended.     

Comparison of several spatial prediction methods for soil pH

A survey of topsoil pH has been designed specifically to compare the performance of several two-dimensional spatial prediction methods. These methods have been classified as global or local, and interpolating or non-interpolating, and smooth or non-smooth predictors. The techniques tested were global means and medians, moving averages, inverse squared distance interpolation, Akima's interpolation, natural neighbour interpolation, quadratic trend surface, Laplacian smoothing splines and ordinary kriging. All methods showed some deficiencies; for example, prediction sums-of-squares were higher than expected for all methods. Interpolating methods either were very poor predictors, or suffered from theoretical drawbacks, or both. Of the non-interpolating methods, Laplacian smoothing splines and kriging generally performed best. Estimates of variance derived from models which assume independent errors were greater than estimates of variance derived from neighbouring pairs of data sites, suggesting that short range correlations definitely exist, and should be taken into account in any predictive modelling.     

土壤质地三角图的规范制作及自动查询

通过分析国内涉及土壤质地三角图的主要参考资料及存在的问题,本文提出了制作规范的土壤质地三角图的基本要求,并新绘制了中文版美国农部制土壤质地三角图。针对人工方式查询土壤质地容易出错以及大批量样品查询效率较低等问题,建立了一种在Excel中通过自定义函数实现土壤质地快速自动的查询方法。     

Associations between soil texture,soil water characteristics and earthworm populations in grassland

Abstract

The aim of the present study was to investigate the relationships between soil physical characteristics and earthworms in a regional-scale field study in Denmark. The earthworm populations along within-field gradients in soil texture were quantified at five field sites, representing dominant soil types of Denmark. Eleven earthworm species were found, but populations were mainly dominated by Aporrectodea tuberculata and A. longa. Despite considerable variation in soil parameters across the five study sites the results suggest that the biomass of anecic worms (or A. longa as a species) was not causally associated with the soil parameters studied. This indicates that there must be other causal factors associated with the abundance (and composition) of anecic worms that are not among the soil texture and structure parameters studied. On the other hand, soil texture (Coarse sand) was associated with the abundance of the dominant endogeic species, A. tuberculata, but not endogeic worms in general. It was hypothesized that anecic and endogeic earthworms might respond to local soil water characteristics rather than soil texture, but this hypothesis could not be confirmed with the present data.     

稻壳掺播对土壤质地及烟叶生产的影响

皖南烟区土壤主要为水稻土,质地偏黏,不利于烟叶生长,为此,在宣城烟区耕作层土壤中掺播一定体积比例的稻壳,探讨其对土壤质地和烟叶生产的影响。结果表明:①随着掺播稻壳比例的增加,0.02～2mm的土壤颗粒和土壤有机质含量增加,土壤体积质量(容重)、含水量和pH降低,但这种趋势随着时间推移而逐渐减缓。②烟田耕作层土壤掺播稻壳总体上有利于烤烟生长,能够提高烟叶产量和品质,经济效益也明显提高。其中以掺播20%～30%的稻壳(稻壳体积/土壤体积)时烤烟的农艺性状、经济性状及工艺评价结果最佳。     

沙质草地生境中植物群落对土壤质地演进的响应

利用空间代替时间,选取6个沙地生境类型(硬质灰钙土H_1,沙质灰钙土H_2,固定沙地H_3,半固定沙地H_4,半流动沙地H_5和流动沙地H_6),通过对其土壤颗粒组成、草地植物群落结构及数量特征的分析,研究沙质草地植物群落演替对土壤沙化过程的响应。结果表明:(1)从硬质灰钙土到流动沙地,物种丰富度分别为9,12,9,9,9,5种;演替初期(H_1-H_2)多年生草本植物种占绝对优势,演替中期(H_3-H_4)一年生植物优势度明显,演替后期(H_5-H_6)则以半灌木占主导地位;各演替阶段禾本科植物均有很高的优势度,在半流动沙地阶段成为最大优势物种,在沙质灰钙土后藜科植物优势度保持相对稳定,菊科和豆科植物分别在固定沙地和半固定沙地阶段逐渐消失。(2)物种多样性在沙质灰钙土最高,到固定沙地持续递减,而到流动沙地时又增大,生态优势度与物种多样性变化趋势正好相反,均匀度表现出先减小后增大的趋势,相邻的两个群落间具有较高的相似度。(3)从硬质灰钙土到流动沙地,0—10cm土层土壤黏粉粒含量极显著减少(p0.01),而粗砂粒极显著增加(p0.01),土壤颗粒分形维数由2.52下降到1.73。(4)一年生草本植物优势度与土壤黏粉粒含量、土壤颗粒分形维数呈显著负相关(p0.05);豆科植物优势度与土壤粉粒组分、土壤颗粒分形维数间呈显著正相关(p0.05);物种多样性与粉粒组分含量呈极显著正相关(p0.01),与细砂粒组分含量、土壤颗粒分形维数间呈显著正相关(p0.05)。     

Validation of topsoil texture derived from agricultural soil maps by current dense soil sampling

Agricultural soil maps were created to facilitate the management of fields. Such maps, at very large scale (1:5000), cover almost the entire agricultural land in Poland. These maps can be very useful for precision field management. However, they were prepared about 40–50 years ago using old mapping techniques with mainly field (organoleptic) examination of soil. For this reason it is necessary to verify agreement between such maps with the current soil status. In the case of detection of disagreements it is important to identify the causes and propose methods for improving soil maps. Only few studies were made in this aspect. The soil texture (ST) of the upper layer, as presented on the agricultural soil maps from the 1960s and 1970s, was compared with the actual ST of the same four fields in northern and central Poland. For laboratory ST determination, soil samples were collected in a dense sampling network. On average, the agreement of the ST presented on the agricultural soil maps and determined in this study was classified as medium for three fields and good for one field. However, the presence of field areas with poor agreement between the ST determined in this study and that shown on the agricultural maps was detected in all investigated fields. Identified causes for this comprised imprecision of ST determination using the organoleptic method, the generalization of the soil quality maps, and erosion processes during the years between soil mapping and this investigation. The improvement of ST presentation on the large scale soil maps might be achieved by quite denser soil sampling for ST analysis supported by ECa, yield and NDVI maps.     

不同取样方式下土壤质地空间插值的精度分析

为研究土壤质地的合理取样方式,进而研究其空间变异情况,为田间施肥及灌溉提供依据,本试验利用地统计学方法和GIS技术,在重庆市彭水县重庆烟草试验站,利用289个表层土样,研究了16 m间距的栅格取样法(对照,253个土样,扣除36个验证样点)、34 m间距的栅格取样法(115个土样)和随机取样法(115个土样)3种取样方式下土壤质地的空间插值精度。3种土壤颗粒指标中粉粒占68.43%,砂粒含量最少,占12.68%,黏粒含量略高于砂粒。砂粒和黏粒具有中等强度的变异性,粉粒具弱变异性,且数据符合正态分布。地统计分析显示,在分析该区域土壤质地时,采用栅格取样方法应适当增大取样间距,而采用随机取样方法可适当缩小取样间距。交叉检验显示,土壤质地成分在3种取样方式下的插值精度均以对照最大,栅格取样次之,随机取样最小。综合考虑插值误差、样品采集和分析成本及时效性等因素,本研究建议在该区域进行土壤质地空间变异规律分析为生产服务时应采用随机取样。