共查询到20条相似文献，搜索用时 593 毫秒
1.
基于地形属性使用回归, 克里格和人工神经网络方法估算饱和导水率空间分布 总被引：6，自引：0，他引：6
Several methods,including stepwise regression,ordinary kriging,cokriging,kriging with external drift,kriging with varying local means,regression-kriging,ordinary artificial neural networks,and kriging combined with artificial neural networks,were compared to predict spatial variation of saturated hydraulic conductivity from environmental covariates.All methods except ordinary kriging allow for inclusion of secondary variables.The secondary spatial information used was terrain attributes including elevation,slope gradient,slope aspect,profile curvature and contour curvature.A multiple jackknifing procedure was used as a validation method.Root mean square error (RMSE) and mean absolute error (MAE) were used as the validation indices,with the mean RMSE and mean MAE used to judge the prediction quality.Prediction performance by ordinary kriging was poor,indicating that prediction of saturated hydraulic conductivity can be improved by incorporating ancillary data such as terrain variables.Kriging combined with artificial neural networks performed best.These prediction models made better use of ancillary information in predicting saturated hydraulic conductivity compared with the competing models.The combination of geostatistical predictors with neural computing techniques offers more capability for incorporating ancillary information in predictive soil mapping.There is great potential for further research and development of hybrid methods for digital soil mapping. 相似文献
2.
模糊c-均值算法在区域土壤预测制图中的应用 总被引：8，自引：2，他引：6
基于模糊c-均值算法和地统计学空间插值,在面积约为1km2的研究区内进行区域土壤预测制图。研究结果表明:根据研究区123个剖面和土钻样点,通过分析它们在形态学上的特征和定量属性,建立了9类诊断特征土层。通过FCM算法模型,获得4类最佳分类数,模糊指数为1.7。类别数目与研究区受地形、母质和土地利用方式影响的主要成土过程决定的土纲下土壤类型数目一致。将经过对称对数比转换的隶属度成分数据进行单一模糊类别隶属度土壤预测制图,4种类别土壤在空间上具有明显的渐变过渡特征,制图结果较理想。在单一类别隶属度土壤图的基础上生成最大隶属度土壤图,与常规土壤调查土壤图具有共同参比的基础。 相似文献
3.
Maren Oelbermann R. Paul Voroney Donald C.L. Kass Andrea M. Schlnvoigt 《Geoderma》2006,130(3-4):356-367
Conversion of forests to agricultural land in the American tropics, through traditional agricultural practices such as shifting cultivation, has not been able to maintain stocks of soil organic carbon (SOC), and increasing population pressure has led to shortened fallow periods, causing further losses of soil fertility. However, land management practices such as agroforestry can provide a sustainable alternative to single cropping because of its ability to maintain or increase the SOC pool. This study quantified SOC and nitrogen (N) pools, gross SOC turnover, residue stabilization efficiency (RSEAC) in the alley crop, soil δ13C partitioning, C3-C abundance and δ15N dynamics in 19- and 10-year Gliricidia sepium and Erythrina poeppigiana alley cropping system. Each system was studied at two fertilizer levels (tree prunings only [−N or −A], and tree prunings plus chicken manure [+N], or Arachis pintoi as a groundcover [+A]), and was compared to a sole crop system. The SOC and N pools were significantly higher (p < 0.05) in the 19-year-old alley crop compared to the sole crop, but not significantly different (p < 0.05) in the 10-year-old system. Soil C and N (%) showed a similar trend as that of the SOC and N pools in both 19- and 10-year-old systems. Gross SOC turnover, to a 20 cm depth, ranged from 12 to 21 years in the 19-year-old alley crop compared to 50 years in the sole crop, and from 20 to 32 years in the 10-year-old alley crop compared to 106 years in the sole crop. The RSEAC ranged from 10% to 58% in the 19-year-old system, and from 3% to 43% in the 10-year-old system. The δ13C signature of the soil shifted significantly (p < 0.05) towards that of C3 vegetation in the alley crop due to the greater input of organic residues from tree prunings compared to the sole crop. The proportion of input from tree prunings only in the 19-year-old alley crop ranged from 14% to 20%, and from 9% to 11% in the 10-year-old system to a soil depth of 20 cm. The δ15N signature of the soil showed two patterns: that of the 19-year-old system being enriched in δ15N, and that of the 10-year-old system being depleted in δ15N compared to the sole crop. The addition of manure in the 19-year-old system has enriched the soil δ15N and in the 10-year-old system the soil was depleted due to the N2-fixing groundcover A. pintoi. 相似文献
4.
《Communications in Soil Science and Plant Analysis》2012,43(12):1452-1464
ABSTRACTSoil texture is a key controlling factor of soil properties and its functions include water and nutrient holding capacity, retention of pollutants, root development, soil biodiversity, and biogeochemical cycling. From the geotechnical standpoint, it is interesting to analyze the soil texture in regions due to its relation with the infiltration and runoff processes and, consequently, the effect on erosion processes. The purpose of this study is to present a methodology that provides the soil texture spatial variation by using Fuzzy logic theory and geostatistical technique in Geographic Information System (GIS) platform. A total of 140 soil samples were taken from topsoil (0–30 cm) in the study area located in the north of Guilan Province, the southern coast of Caspian Sea, Northern Iran. The soil textural classes were converted to numerical values (fuzzy values) using the fuzzy logic concept. The fuzzy values were spatially interpolated by ordinary kriging method such that the fitted model on experimental semi-variogram was exponential with moderate structure. The results showed the accuracy of soil texture predictive map was acceptable according to the values of normalized root-mean-square error for train data set (0.182) and test data set (0.179). The knowledge of the spatial variability of soil properties such as the soil texture can be an important tool for land-use planning in order to reduce the potential soil losses during rainy seasons. The results indicated that the integration of fuzzy logic, geostatistics, and GIS can improve the interpolation process. 相似文献
5.
There is limited information on the effects of tillage practices on soil hydraulic properties, especially changes with time. The objective of this study was to evaluate on a long-term field experiment the influence of conventional tillage (CT), reduced tillage (RT) and no-tillage (NT) on the dynamics of soil hydraulic properties over 3 consecutive 16–18 month fallow periods. Surface measurements of soil dry bulk density (ρb), soil hydraulic conductivity (K(ψ)) at −14, −4, −1 and 0 cm pressure heads using a tension disc infiltrometer, and derived hydraulic parameters (pore size, number of pores per unit of area and water-transmission porosity) calculated using the Poiseuille's Law were taken on four different dates over the fallow period, namely, before and immediately after primary tillage, after post-tillage rains and at the end of fallow. Under consolidated structured soil conditions, NT plots presented the most compacted topsoil layer when compared with CT and RT. Soil hydraulic conductivity under NT was, for the entire range of pressure head applied, significantly lower (P < 0.05) than that measured for CT and RT. However, NT showed the largest mean macropore size (0.99, 0.95 and 2.08 mm for CT, RT and NT, respectively; P < 0.05) but the significantly lowest number of water-conducting pores per unit area (74.1, 118.5 and 1.4 macropores per m2 for CT, RT and NT, respectively; P < 0.05). Overall, water flow was mainly regulated by macropores even though they represented a small fraction of total soil porosity. No significant differences in hydraulic properties were found between CT and RT. In the short term, tillage operations significantly increased K (P < 0.05) for the entire range of pressure head applied, which was likely a result of an increase in water-conducting mesopores despite a decrease in estimated mesopore diameter. Soil reconsolidation following post-tillage rains reduced K at a rate that increased with the intensity of the rainfall events. 相似文献
6.
The main objectives of this study were to model the relationship between WRB-1998 soil groups and terrain attributes and predict the spatial distribution of the soil groups using digital terrain analysis and multinomial logistic regression integrated in GIS in the Vestfold County of south-eastern Norway. A digital elevation model of 25 meter grid resolution was used to derive fifteen terrain attributes. A digitized soil map of thirteen WRB soil groups at the scale of 1:25,000 was used to obtain the reference soil data for model building and validation. First, the relationships between the soil groups and the terrain attributes were modeled using multinomial logistic regression. Then, the probability that a given soil type is present at a given pixel was determined from the logit models in ARCGIS to continuously map each soil group's spatial distribution. Elevation, flow length, duration of daily direct solar radiation, slope, aspect and topographic wetness index were found to be the most significant terrain attributes correlating with the spatial distribution of the soil groups. The prediction showed higher mean probability values for each soil group in the areas actually covered by that soil group compared to other areas, indicating the reliability of the prediction. However, the prediction performed poorly for soil groups that are not greatly influenced by topography but by other factors such as human activities. 相似文献
7.
The particle density of soil (ρS) represents one of the soil's basic physical properties and it depends on the composition of both the mineral and the organic soil components. It therefore varies for different soils, e.g. within the group of mineral soils, and ranges from 2.4–2.9 g cm−3. Hence, awareness of this variability is important for properties estimated by a calculation involving particle density. Because ρS depends on both the soil's solid mineral particles and soil organic matter composition, we derived a function based on the mixture ratio of these two soil components. This approach represents a further development of earlier investigations dealing with the influence of organic carbon (Corg) on ρS. To parameterise this function, two data sets were used: (1) data from soils with Corg contents between 0% and 54.88% and corresponding values of ρS between 1.49 and 2.72 g cm−3; and (2) data from soils of 17 German long-term experiments contrasting in soil texture and in soil mineral inventory. Data set 1 was used to quantify the influence of soil organic matter on ρS, and data set 2 was used to calculate the influence of mineral matrix on ρS. The soil organic matter has two major influences on ρS: (1) via a mass effect (expressed as a mixture ratio between organic and mineral soil components); and (2) via a quality effect (expressed as calculated changes in particle density of organic soil components). Here, we calculated that with increasing content of soil organic matter (0–100%), the particle density of organic soil components rose from about 1.10 to 1.50 g cm−3, and present possible reasons for this phenomenon. Additionally, we demonstrate that the mineral matrix of the soil affects ρS especially via variations in the mineral inventory, but conclude that differences in particle size distribution of soils were to a lesser extent suitable for describing the influence of the mineral matrix on ρS. Overall, using our approach should generate more realistic values of ρS, and consequently of all calculated parameters which are sensitive to ρS. 相似文献
8.
9.
基于不同地表曲面模型预测土壤有机碳含量 总被引：1，自引：0，他引：1
Local terrain attributes,which are derived directly from the digital elevation model,have been widely applied in digital soil mapping.This study aimed to evaluate the mapping accuracy of soil organic carbon (SOC) concentration in 2 zones of the Heihe River in China,by combining prediction methods with local terrain attributes derived from different polynomial models.The prediction accuracy was used as a benchmark for those who may be more concerned with how accurately the variability of soil properties is modeled in practice,rather than how morphometric variables and their geomorphologic interpretations are understood and calculated.In this study,2 neighborhood types (square and circular) and 6 representative algorithms (Evans-Young,Horn,Zevenbergen-Thorne,Shary,Shi,and Florinsky algorithms) were applied.In general,35 combinations of first-and second-order derivatives were produced as candidate predictors for soil mapping using two mapping methods (i.e.,kriging with an external drift and geographically weighted regression).The results showed that appropriate local terrain attribute algorithms could better capture the spatial variation of SOC concentration in a region where soil properties are strongly influenced by the topography.Among the different combinations of first-and second-order derivatives used,there was a best combination with a more accurate estimate.For different prediction methods,the relative improvement in the two zones varied between 0.30％ and 9.68％.The SOC maps resulting from the higher-order algorithms (Zevenbergen-Thorne and Florinsky) yielded less interpolation errors.Therefore,it was concluded that the performance of predictive methods,which incorporated auxiliary variables,could be improved by attempting different terrain analysis algorithms. 相似文献
10.
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。 相似文献
11.
基于土壤系统分类的土壤类型和界线确定研究——以句容大顶山研究区为例 总被引：1，自引：0，他引：1
我国土壤系统分类的土壤调查技术仍处于研究阶段,借鉴传统的发生分类的土壤调查方法,本文主要研究了基于土壤系统分类的土壤类型与界线确定。土壤类型是通过常规布点法进行主要剖面点位置的设定,并从地形分级和制图精度方面分析确定其数量,对挖掘的土壤剖面利用诊断层和诊断特性定名得出。土壤界线则是利用内插法在调查路线上确定土壤类型分界点后,结合遥感图像将不同调查路线相同土壤类型分界点用平滑曲线连接最终确定。研究表明,土壤类型和土壤界线与母质、地形、植被和土地利用方式等多种景观因素有密切的联系,在土壤调查中要充分考虑景观因素的影响。 相似文献
12.
Over the last two decades, soil cultivation practices in the southern Argentinean Pampas have been changing from a 7 year cash-crop production system alternated with 2–3 years under pasture, to a continuous cropping system. A better understanding of the impact of the period of time a field has been under continuous cropping on a broad spectrum of soil properties related to soil quality is needed to target for sustainable cropping systems. The objectives of this study were to: (i) assess the relationship between physical and chemical soil parameters related to soil quality and (ii) identify soil quality indicators sensitive to soil changes under continuous cropping systems in the Argentinean Pampas.
Correlation analysis of the 29 soil attributes representing soil physical and chemical properties (independent variables) and years of continuous cropping (dependent variable) resulted in a significant correlation (p < 0.05) in 78 of the 420 soil attribute pairs. We detected a clear relationship between hydraulic conductivity at tension h (Kh) and structural porosity (ρe); ρe being a simple tool for monitoring soil hydraulic conditions.
Soil tillage practice (till or no-till) affected most of the soil parameters measured in our study. It was not possible to find only one indicator related to the years under continuous cropping regardless of the cultivation practice. We observed a significant relationship between years under continuous cropping and Kh under no-till (NT) and wheat fallow (p < 0.001, R2 = 0.70). Under these conditions, K−40 diminished as the number of years under continuous cropping increased.
The change in mean weight diameter (CMWD) was the only physical parameter related to the number of years under continuous cropping, explaining 36% of the variability in the number of years under continuous cropping (p < 0.001) The combination of three soil quality indicators (CMWD, partial R2 = 0.38; slope of the soil water retention curve at its inflexion point (S), partial R2 = 0.14 and cation exchange capacity (CEC), partial R2 = 0.13) was able to explain, in part, the years under continuous cropping (R2 = 0.65; p value > 0.001), a measure related to soil quality. 相似文献
13.
14.
土壤质地空间变异对于区域土壤水分和溶质循环模拟研究至关重要。基于转移概率(Markov链)的随机条件模拟是一种能融合多源信息技术的地统计学模拟方法,与传统插值法和基于变异函数的条件模拟相比有诸多优势。本研究在封丘县域范围内采集了102个土壤剖面共计1 020个土壤样品,运用转移概率的地质统计学方法构建了土壤质地种类的三维空间分布模型,在500次随机模拟的基础上,分别得到了土壤质地种类空间分布图及其概率分布图。结果表明,一维垂向转移概率可以很好地描述研究区垂直方向上各类型土壤空间分布的不对称性与各向异性。采用转移概率的方式来描述土壤质地种类空间分布的不确定性,能够有效克服传统地统计学基于交叉变异函数或空间插值建模的缺点。 相似文献
15.
基于超高空间分辩率无人机影像的面向对象土地利用分类方法 总被引：2，自引：2，他引：0
为明确基于无人机超高空间分辨率影像的土地利用分类方法,尤其是有效特征和算法的选择,该研究获取吉林省德惠市一农耕区超高分无人机影像,获取区域正射影像图和数字表面模型,计算地形指标,采用面向对象方法进行土地利用分类研究。首先,采用随机森林算法,以光谱特征为基础,依次引入指数、形态、地形、纹理特征,建立5种特征选择方案,分析各类特征对分类效果的影响。其次,以Boruta特征选择算法获取的优化特征集为基础,采用随机森林算法、朴素贝叶斯算法、逻辑回归算法和支持向量机算法分类,分析不同算法的分类效果。结果表明:采用5种特征选择方案分类,引入形态特征时总体精度降低,引入其他特征时总体精度逐渐提高。5种特征共同参与的分类效果最佳,总体精度为98.04%,Kappa系数为0.980。错分主要发生在裸地和宅基地,漏分主要发生在草地、裸地、水渠和道路。错分和漏分主要是因为这几种类型对象具有相似的光谱、形态、纹理特征或相似的分布位置。采用优化特征集分类时,相比其他算法,随机森林算法更擅长处理高维特征集,获得最高的总体精度98.19%,最低的错分和漏分误差,分类效果最佳。借助无人机超高空间分辨率影像提取地形信息、形态信息,可以有效辅助土地利用分类,并能提高传统分类方法精度。 相似文献
16.
Assessment of soil texture class on agricultural fields using ECa,Amber NDVI,and topographic properties 总被引：1，自引：0，他引：1 
下载免费PDF全文
下载免费PDF全文 Michał Stępień Stanisław Samborski Dariusz Gozdowski Eike Stefan Dobers Jarosław Chormański Jan Szatyłowicz 《植物养料与土壤学杂志》2015,178(3):523-536
Soil texture (ST) is relatively stable over time, although it may change due to erosion, clay eluviation, and other processes. Soil texture affects soil quality, productivity and management. Therefore, indirect, accurate methods for assessing of soil texture classes (STCs) are needed in agricultural practice. A study was performed on four production fields in northern and central Poland to compare the fitting performance of STC models based on apparent electrical conductivity (ECa), topographic properties (elevation, slope gradient and wetness index) and Amber NDVI measurements. One common and accurate indicator of STCs was not found for all study fields. On average, ECa was most accurate in indicating areas of different STCs within the fields, but it tended to overestimate the size of sandy areas on loamy fields and vice versa. The accuracy of STC assessment using ECa measurements may be biased due to imperfect soil drainage, high elevations, which increase evaporation and STC variation with depth. STC assessment using Amber NDVI measurements may be useful, particularly on flat and sandy fields, but the results are affected by the same factors as ECa, and additionally by crop growth stages and by the weather conditions in the period preceding the measurements. Despite the good quantitative results of the STC assessment by elevation (one field) and by the topographic wetness index (another field), both terrain attributes failed to accurately indicate the distribution of some STC areas within each field. Therefore, in landscapes developed from deposits of the last glaciation relevant ST differences might not sufficiently be detected by the analysis of terrain attributes alone. The selection of STC predictors and evaluation of the assessment quality must consider both the quantitative indicators such as correlation and determination coefficients describing relationships between ST and ECa, NDVI and topography and percentage of a field area with accurately indicated STC and the distribution of areas with different STCs within a field. The use of ECa, NDVI values, and topographic properties for STC assessment may be useful in reducing costs of soil sampling and analysis, but cannot replace it. 相似文献
17.
Inspired by the catena concept, we returned to the landscape of origin for this idea in central Uganda and applied quantitative soil-landscape-modeling techniques to the study of soil formation. In particular, we employed Generalized Least Squares (GLS) regression to identify potential terrain (and other) controls on surface texture with the goal of explaining texture contrast formation for well-drained red soils, and fluvial deposition for yellow-gray lowland soils. Well-drained red soils and poorly drained yellow-gray soils were identified using a high-resolution spectroradiometer and soil redness was highly correlated with a measure of elevation above the valley floor. For the yellow-gray soils, sand content and sand grain size increased with greater slope gradient and in converging areas—evidence supporting fluvial control of surface soil composition. For red soils, texture contrast increased on lower backslopes, decreased over ferricrete-rich parent material, and had no significant association with either infiltration- or runoff-influenced locations. Surprisingly, texture contrast was also reduced or even inverted on fine scale convexities (<18 m in diameter). These findings were consistent with the following theoretical processes: (1) sand deposition on lower-backslopes, increasing texture contrast; (2) ferricrete weathering at the soil surface, reducing texture contrast; and (3) texture contrast inversion through termite turbation. A number of techniques proved useful in this study: (a) the use of GLS regression supported the interpretation of model coefficients in an explicitly spatial context; (b) thin-plate splines provided a flexible means of controlling the spatial scale of terrain index computation; and (c) the development of a new downslope-looking Elevation Above Stream (EASy) terrain index provided a powerful tool for the identification of poorly drained soils—a tool that could have applications in a variety of landscapes and environments. Overall, we found that the catena concept as a soil-terrain-modeling approach worked best on lower, less well-drained, depositional landscape components. For the well-drained portion of this landscape, there was little evidence to suggest that soil formation was controlled by hillslope hydrology as proposed by the catena concept. The catena concept interpreted in a broad sense challenges pedologists to explicitly link soil geographic patterns with specific soil forming processes and environmental controls—an approach that proved valuable in this study. 相似文献
18.
Seasonal variations in soil erosion resistance during concentrated flow for a loess-derived soil under two contrasting tillage practices 总被引：2，自引：0，他引：2
Soil erodibilty during concentrated flow (Kc) and critical flow shear stress (τcr), both reflecting the soil's resistance to erosion by concentrated runoff, are important input parameters in many physically-based soil erosion models. Field data on the spatial and temporal variability of these parameters is limited but crucial for accurate prediction of soil loss by rill or gully erosion. In this study, the temporal variations in Kc and τcr for a winter wheat field on a silt loam soil under three different tillage practices (conventional ploughing, CP; shallow non-inversion tillage, ST; deep non-inversion tillage, DT) in the Belgian Loess Belt were monitored during one growing season. Undisturbed topsoil samples (0.003 m3) were taken every three weeks and subjected to five different flow shear stresses (τ = 4–45 Pa) in a laboratory flume to simulate soil detachment by concentrated flow. To explain the observed variation, relevant soil and environmental parameters were measured at the time of sampling. Results indicated that after two years of conservation tillage, Kc(CP) > Kc(DT) > Kc(ST). Kc values can be up to 10 times smaller for ST compared to CP but differences strongly vary over time, with an increasing difference with decreasing soil moisture content. The beneficial effects of no-tillage are not reflected in τcr. Kc values vary from 0.006 to 0.05 sm−1 for CP and from 0.0008 to 0.01 sm−1 for ST over time. Temporal variations in Kc can be mainly explained by variations in soil moisture content but consolidation effects, root growth, residue decomposition and the presence of microbiotic soil crusts as well play a role. τcr values increase with increasing soil shear strength but Kc seems more appropriate to represent the temporal variability in soil erosion resistance during concentrated flow. The large intra-seasonal variations in Kc, which are shown to be at least equally important as differences between different soil types reported in literature, demonstrate the importance of incorporating temporal variability in soil erosion resistance when modelling soil erosion by concentrated flow. 相似文献
19.
A comparison of kriging, cubic splines and classification for predicting soil properties from sample information 总被引：9，自引：0，他引：9
The clay content of the topsoil in two regions of contrasting physiography was predicted from sample data using four different procedures. The predictors were the means of mapped classes, the usual kriging estimator, a cubic spline interpolator and a kriging estimator within classes using a pooled within-class variogram. The performances of the procedures were evaluated and compared. In the first region, Sandford St Martin on Jurassic sediments where there were some abrupt changes in soil, the classification predicted best within those classes bounded by sharp change. Elsewhere the usual kriging performed somewhat better, and kriging within classes was still more precise. In the second region, Yenne on the alluvial plain of the Rhone where the soil varied gradually, kriging performed better than classification, though a small improvement resulted from combining kriging with classification. Both prediction by class means and kriging attempt to minimize the estimation variance, and their mean prediction variances were close to the theoretical values overall. Spline interpolation is more empirical, and though it followed the abrupt changes better than kriging, it fluctuated excessively elsewhere, and its overall performance was poorer than that of kriging. 相似文献
20.
There is increasing interest in developing automatic procedures to segment landscapes into soil spatial entities that replace conventional, expensive manual procedures for delineating and classifying soils. Geographic object-based image analysis (GEOBIA) partitions remote sensing imagery or digital elevation models into homogeneous image objects based on image segmentation. We used an object-based methodology for the detailed delineation and classification of soil types using digital maps of topography and vegetation as soil covariates, based on the Random Forests (RF) classifier. We compared the object-based method's results with those of a pixel-based classification using the same classifier. We used 18 digital elevation model derivatives and 5 remote sensing indices that were related to vegetation cover and soil. Using 171 soil profiles with their associated environmental variable values, the RF method was used to identify the most important soil type predictors for use in the segmentation process. A stack of raster-geodatasets corresponding to the selected predictors was segmented using a multi-resolution segmentation algorithm, which resulted in homogeneous objects related to soil types. These objects were further classified as soil types using the same method, RF. We also conducted a pixel-based classification using the same classifier and soil profiles, and the resulting maps were assessed in terms of their accuracy using 30% of the soil profiles for validation. We found that GEOBIA was an effective method for soil type mapping, and was superior to the pixel-based approach. The optimized object-based soil map had an overall accuracy of 58%, which was 10% higher than that of the optimized pixel-based map. 相似文献