我国黑土区受自然因素和人为因素共同影响，土壤有机质含量持续下降，严重威胁国家粮食安全。精准解析土壤有机质空间分布规律对耕地保护具有战略意义，现有常规地统计学采样方法存在空间覆盖不均和样本代表性不足的缺陷，影响了土壤有机质制图精度。基于此，本研究以科尔沁左翼中旗为研究区，以有机质样点数据为例，兼顾样点的空间均匀性和代表性，研究了一种基于权重调整的土壤有机质空间制图方法，主要内容和结果如下：

（1）土壤有机质样点数据探索性分析和样点类型划分。对研究区内土壤有机质的各项空间特征进行描述，从不同角度剖析土壤有机质的分布规律，包括描述性特征统计、空间自相关、全局趋势分析和空间异质性分析；并结合土壤有机质样点在地理空间和特征空间的分布，研究了一种样点类型划分方法。结果表明：研究区有机质含量为中等变异程度，有机质样点基本符合正态分布，在空间上呈现极强的正相关性，全局趋势模型为二阶多项式模型，最优半变异函数拟合模型为球状模型。样点在地理空间可分为聚集样点、均匀样点和稀疏样点3类，在特征空间可分为高代表性样点和低代表性样点2类。

（2）研究基于权重调整的有机质空间制图方法。基于四分位法和局部Moran's I确定离群样点，针对离群样点的不同类型研究基于地理空间和特征空间的权重调整方案，设置3种对比方案评估不同权重调整方案对空间制图精度的影响，采用RMSE、MAE、AC和R²定量评价不同方案的制图精度，根据误差分布和空间分布分析权重调整不同样点类型对制图结果的影响，最后评估基于地理空间和特征空间的权重调整方法制图的准确性。结果表明：基于地理空间和特征空间的权重调整方法在权重调整样点数最少的情况下显著提高了制图精度，相较于原始样点空间制图结果，其RMSE减少了0.424 g/kg，MAE减少了0.311 g/kg，AC提高了0.072，R²提高了0.117，该方法误差较大的区域面积最少，制图结果能更加清楚地体现有机质含量的高低差异，与现有数据也有较高的一致性。

（3）基于权重调整的空间制图方法与常规空间制图方法对比分析。选取反距离加权和随机森林两种模型与基于权重调整的有机质空间制图方法从制图精度、误差分布及空间分布特征进行对比分析，深入分析基于权重调整的有机质空间制图方法的优势。对比结果表明：基于权重调整的有机质空间制图方法RMSE、MAE、AC和R²均优于常规空间制图方法；在误差控制方面表现最优，绝对误差范围较小，具有更高的空间精度；有机质空间分布高值和低值的分界线更明显，较好地反映了土壤有机质的空间异质性。该方法可为区域土壤属性制图提供技术支撑，对区域土壤变异性研究具有重要意义。

Under the combined influence of natural and anthropogenic factors, the soil organic matter content in China's black soil region continues to decline, posing a serious threat to national food security. Accurately analyzing the spatial distribution patterns of soil organic matter is of strategic significance for cultivated land protection. However, conventional geostatistical sampling methods suffer from uneven spatial coverage and insufficient sample representativeness, which compromise the accuracy of soil organic matter mapping. To address these limitations, this study selected Horqin Left Wing Middle Banner as the study area and utilized soil organic matter sample data to develop a weight-adjusted spatial mapping method for soil organic matter that balances spatial uniformity and representativeness. The main contents and results are as follows:

(1) An exploratory analysis of soil organic matter sample point data was conducted, and the sample point types were classified. The spatial characteristics of soil organic matter in the study area were described. The distribution of soil organic matter was analyzed from different perspectives, including descriptive statistics, spatial autocorrelation, global trend analysis, and spatial heterogeneity analysis. A method for dividing sample point types was investigated by combining the distribution of soil organic matter sample points in geographic and feature spaces. The results showed that the organic matter content in the study area was moderately variable; the organic matter sample points generally conformed to a normal distribution; the organic matter sample points exhibited strong positive spatial correlation; the global trend model was a second-order polynomial; and the optimal semi-variance function fitting model was spherical. Sample points can be categorized into three types in geographic space: aggregated, uniform, and sparse. In feature space, sample points can be categorized into two types: high and low representative.

(2) Research organic matter spatial mapping methods based on weight adjustment. Determine outlier sample points using the quartile method and Local Moran's I. Study weight adjustment schemes based on geospatial and feature space for different types of outlier sample points. Set up three comparison schemes to evaluate the impact of different weight adjustment schemes on spatial mapping accuracy. Quantitatively evaluate the mapping accuracy of the different schemes using RMSE, MAE, AC, and R². Analyze the impact of different types of points on the mapping results based on error distribution and spatial weight adjustment distribution. Finally, evaluate the mapping accuracy of the geospatial and feature-space method. Analyze the effects of different types of points on mapping results according to error and spatial distributions. Finally, evaluate the mapping accuracy of geospatial and feature-space-based weight adjustment methods. The results show that the geospatial and feature-space based weight adjustment method significantly improves mapping accuracy using the minimum number of weight adjustment sample points. Compared to the original sample point mapping results, the RMSE is reduced by 0.424 g/kg, the MAE is reduced by 0.311 g/kg, and the AC is improved by 0.072 and R² by 0.117. The area with large errors was minimized, and the mapping results could more clearly reflect differences in high and low levels of organic matter content. There was also a high degree of consistency with the available data.

(3) A comparative analysis of the weight-adjusted spatial mapping method and the conventional spatial mapping method is conducted. Two models, inverse distance weighting and random forest, were selected for comparison with the organic matter spatial mapping method based on weight adjustment. This comparison was made in terms of mapping accuracy, error distribution, and spatial distribution characteristics. The goal was to deeply analyze the advantages of the organic matter spatial mapping method based on weight adjustment. The results of the comparison show that the RMSE, MAE, AC, and R² of the weight-adjusted organic matter spatial mapping method are better than those of the conventional method. The weight-adjusted method optimally controls error, with smaller absolute error ranges and higher spatial accuracy. The dividing line between high and low organic matter values is clearer, reflecting the spatial heterogeneity of soil organic matter. This method can provide technical support for mapping regional soil properties and is significant for studying regional soil variability.

[1] 仇少君, 郝小雨, 赵士诚, 等. 黑土地肥力下降成因与保护利用对策——基于养分资源高效利用视角分析[J]. 植物营养与肥料学报, 2025, 31(1): 1-11.

[2] Zhao X, Zhang Z, Zhao M, et al. The quantified and major influencing factors on spatial distribution of soil organic matter in provincial-scale farmland-a case study of shandong province in Eastern China[J]. Applied Sciences, 2023, 13(6): 3738.

[3] 刘星雨, 谢雷雷, 刘刚, 等. 典型黑土区杨树水土保持林带对耕地土壤抗侵蚀能力的影响[J]. 生态学报, 2024, 44(24): 11286-11294.

[4] 张娜丽, 澹腾辉, 郭忠录, 等. 东北黑土带典型农田土壤质量退化评价[J]. 水土保持研究, 2025, 32(3): 130-139.

[5] 鲍丽然, 周皎, 李瑜, 等. 渝西北土壤有机质空间变异及影响因素分析[J]. 西南农业学报, 2017, 30(11): 2541-2547.

[6] Liu J, Hong Y, Hu B, et al. Hyperspectral inversion of soil organic matter based on improved ensemble learning method[J]. Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, 2025: 126302.

[7] 王丽萍, 刘焕军, 郑树峰, 等. 东北农牧交错带耕地土壤有机质遥感反演研究[J]. 土壤, 2022, 54(1): 184-190.

[8] 张霄羽, 姚艳敏, 颜祥照. 中红外光谱土壤有机质含量估测研究进展[J]. 中国土壤与肥料, 2021, 4: 327-336.

[9] 曹佳萍, 张黎明, 邱龙霞, 等. 基于稀疏样点的南方丘陵地区耕地土壤有效磷制图[J]. 中国生态农业学报, 2022, 30(2): 290-301.

[10] Wang F, Zhang P, Chen S, et al. Spatial prediction of soil texture in low-relief agricultural areas using rice and wheat growth information with spatiotemporal stability[J]. Remote Sensing, 2025.

[11] 石淑芹, 曹祺文, 李正国, 等. 区域尺度土壤养分的协同克里格与普通克里格估值研究[J]. 干旱区资源与环境, 2014, 28(5): 109-114.

[12] 陈书林, 刘元波, 温作民. 卫星遥感反演土壤水分研究综述[J]. 地球科学进展, 2012, 27(11): 1192-1203.

[13] 魏亚楠. 山东省不同地形耕地土壤有机质遥感反演研究[D]. 泰安:山东农业大学, 2023.

[14] 张静静, 周卫红, 邹萌萌, 等. 高光谱遥感监测大面积土壤重金属污染的研究现状、原理及展望[J]. 江苏农业科学, 2018, 46(12): 9-13.

[15] 贾鲁净, 杨联安, 冀泳帆, 等. 卫星遥感反演土壤有机质研究进展[J]. 遥感信息, 2023, 38(2): 1-9.

[16] 刘焕军, 赵春江, 王纪华, 等. 黑土典型区土壤有机质遥感反演[J]. 农业工程学报, 2011, 27(8): 211-215.

[17] 张法升, 曲威, 尹光华, 等. 基于多光谱遥感影像的表层土壤有机质空间格局反演[J]. 应用生态学报, 2010, 21(4): 883-888.

[18] 张淑杰, 朱阿兴, 刘京, 等. 基于样点的数字土壤属性制图方法及样点设计综述[J]. 土壤, 2012, 44(6): 917-923.

[19] 朱阿兴, 杨琳, 樊乃卿, 等. 数字土壤制图研究综述与展望[J]. 地理科学进展, 2018, 37(1): 66-78.

[20] 陈浩峰, 方彦奇, 赵国凤, 等. 黄棕壤性水稻土有机质含量高光谱反演研究[J]. 江西农业学报, 2024, 36(4): 59-66.

[21] 陈力菡. 基于阈值——指数提取和辐射传输模型的高光谱土壤重金属与有机质反演[D]. 徐州: 中国矿业大学, 2024.

[22] 李莹莹, 赵正勇, 杨旗. 数字土壤制图在土壤养分方面的研究综述[J]. 江西农业学报, 2021, 33(7): 61-67.

[23] 杜霞飞, 宗良纲, 张琪惠, 等. 基于空间差异的有机茶园环境适宜性评价的土壤采样方法[J]. 农业环境科学学报, 2017, 36(1): 150-157.

[24] 王彬武, 周卫军, 马苏, 等. 耕地土壤速效钾含量的空间预测方法研究[J]. 地理与地理信息科学, 2011, 27(5): 91-94+2.

[25] Martin S A, Manucharyan G E, Klein P. Synthesizing sea surface temperature and satellite altimetry observations using deep learning improves the accuracy and resolution of gridded sea surface height anomalies[J]. Journal of Advances in Modeling Earth Systems, 2023, 15(5): e2022MS003589.

[26] 陈娟, 罗明良, 张斌, 等. 反距离加权插值参数变化对元谋冲沟DEM构建的影响[J]. 中国水土保持科学, 2015, 13(1): 29-34.

[27] 易桂花, 张廷斌, 何奕萱, 等. 四种气温空间插值方法适用性分析[J]. 成都理工大学学报(自然科学版), 2020, 47(1): 115-128.

[28] 咸阳, 宋江辉, 王金刚, 等. 基于环境变量筛选与机器学习的土壤养分含量空间插值研究[J]. 农业机械学报, 2024, 55(10): 379-391.

[29] 孙英君, 王劲峰, 柏延臣. 地统计学方法进展研究[J]. 地球科学进展, 2004, (2): 268-274.

[30] 马海建, 秦其明, 金川. AVHRR 1B定位数据经验插值模型的误差分析[J]. 地理与地理信息科学, 2007, (5): 19-21.

[31] 肖斌, 赵鹏大, 侯景儒. 地质统计学新进展[J]. 地球科学进展, 2000, (3): 293-296.

[32] 朱会义, 刘述林, 贾绍凤. 自然地理要素空间插值的几个问题[J]. 地理研究, 2004, 23(4): 425-432.

[33] 马千里. 典型黑土区高空间分辨率耕地土壤有机质遥感反演模型[D]. 长春: 中国科学院大学, 2024.

[34] 李新, 程国栋, 卢玲. 空间内插方法比较[J]. 地球科学进展, 2000, (3): 260-265.

[35] 张海平, 周星星, 代文. 空间插值方法的适用性分析初探[J]. 地理与地理信息科学, 2017, 33(6): 14-18+105.

[36] 张磊. 不确定性指导下的自适应数字土壤制图补样方法[D]. 南京: 南京师范大学, 2018.

[37] 王会肖, 张超. 利用MATLAB研究土壤水分空间变异初探[J]. 中国生态农业学报, 2007, (1): 127-130.

[38] 赫晓慧, 温仲明. 小流域地形因子影响下的土壤水分空间变异性研究[J]. 水土保持研究, 2008, (2): 80-83+7.

[39] 庞国伟, 杨勤科, 张爱国, 等. 陕西省水蚀土壤因子指标插值方法比较研究[J]. 水土保持通报, 2009, 29(2): 176-181+215+182.

[40] 张金兰, 欧阳婷萍, 黄铁兰, 等. 农田表层土壤镉的典型空间插值方法对比研究[J]. 生态科学, 2017, 36(6): 130-136.

[41] 赵永存, 史学正, 于东升, 等. 不同方法预测河北省土壤有机碳密度空间分布特征的研究[J]. 土壤学报, 2005, (3): 379-385.

[42] Mishra U, Lal R, Liu D, et al. Predicting the spatial variation of the soil organic carbon pool at a regional scale[J]. Soil Science Society of America Journal, 2010, 74(3): 906-914.

[43] 孙孝林, 王会利, 宁源. 样点代表性等级采样法在丘陵山区土壤表层有机质制图中的应用[J]. 土壤, 2014, 46(3): 439-445.

[44] Zhu A X. Mapping soil landscape as spatial continua: the neural network approach[J]. Water Resources Research, 2000, 36(3): 663-677.

[45] Rahman F, Khan M A R, Alam M. Hybrid PSO-GA optimization for enhancing decision tree performance in soil classification and crop cultivation prediction[J]. Evolutionary Intelligence, 2025, 18(1): 30.

[46] Behrens T, Forster H, Scholten T, et al. Digital soil mapping using artificial neural networks[J]. Journal of plant nutrition and soil science, 2005, 168(1): 21-33.

[47] Selige T, Bohner J, Schmidhalter U. High resolution topsoil mapping using hyperspectral image and field data in multivariate regression modeling procedures[J]. Geoderma, 2006, 136(1-2): 235-244.

[48] 周斌, 王繁. 基于决策树模型的土壤性质空间推断[J]. 土壤通报, 2004, (4): 385-390.

[49] Wei L, Yuan Z, Wang Z, et al. Hyperspectral inversion of soil organic matter content based on a combined spectral index model[J]. Sensors, 2020, 20(10): 2777.

[50] 陆忠军, 刘克宝, 辛蕊, 等. 基于Sentinel-2影像的青冈县土壤表层有机质空间格局反演[J]. 黑龙江农业科学, 2022, (10): 89-93.

[51] Mutuo P K, Shepherd K D, Albrecht A, et al. Prediction of carbon mineralization rates from different soil physical fractions using diffuse reflectance spectroscopy[J]. Soil Biology and Biochemistry, 2006, 38(7): 1658-1664.

[52] Tan Y, Jiang Q, Yu L, et al. Reducing the moisture effect and improving the prediction of soil organic matter with VIS-NIR spectroscopy in black soil area[J]. IEEE Access, 2021, 9: 5895-5905.

[53] 胡宇. 达拉特旗黄河南岸盐碱化土壤水盐及有机质遥感反演研究[D]. 呼和浩特:内蒙古农业大学, 2024.

[54] Lin L, Liu X. Soil-moisture-index spectrum reconstruction improves partial least squares regression of spectral analysis of soil organic carbon[J]. Precision Agriculture, 2022, 23(5): 1707-1719.

[55] Cao X, Li X, Ren W, et al. Hyperspectral estimation of soil organic matter content using grey relational local regression model[J]. Grey Systems: Theory and Application, 2021, 11(4): 707-722.

[56] Chen F, Kissel D E, West L T, et al. Field‐scale mapping of surface soil organic carbon using remotely sensed imagery[J]. Soil Science Society of America Journal, 2000, 64(2): 746-753.

[57] 宋金红, 吴景贵, 赵欣宇, 等. 基于TM数据的黑土有机质含量空间格局反演研究[J]. 土壤学报, 2015, 52(6): 1422-1429.

[58] Ahmed Z, Iqbal J. Evaluation of Landsat TM5 multispectral data for automated mapping of surface soil texture and organic matter in GIS[J]. European Journal of Remote Sensing, 2014, 47(1): 557-573.

[59] 王琼, 陈兵, 戴建国, 等. Landsat8 与国产卫星数据在表层土壤有机质反演中的对比研究[J]. 绿洲农业科学与工程, 2017, 3(3): 11-20.

[60] 王琼, 陈兵, 王方永, 等. 基于HJ卫星的棉田土壤有机质空间分布格局反演[J]. 农业工程学报, 2016, 32(1): 174-180.

[61] Medhioub E, Bouaziz M, Bouaziz S. Spatial estimation of soil organic matter content using remote sensing data in southern Tunisia[C]//Advances in Remote Sensing and Geo Informatics Applications: Proceedings of the 1st Springer Conference of the Arabian Journal of Geosciences (CAJG-1), Tunisia 2018. Springer International Publishing, 2019: 215-217.

[62] Zhai M. Inversion of organic matter content in wetland soil based on Landsat 8 remote sensing image[J]. Journal of Visual Communication and Image Representation, 2019, 64: 102645.

[63] 陈超群, 戴慧敏, 冯雨林, 等. 基于Sentinel-2A的孙吴地区土壤有机质反演研究[J]. 物探与化探, 2022, 46(5): 1141-1148.

[64] 孙欣琪, 张蚌蚌, 柴朝卿, 等. 沙地整治下榆林土地利用及土壤有机质时空分异特征[J]. 农业工程学报, 2022, 38(24): 207-217.

[65] Nowkandeh S M, Noroozi A A, Homaee M. Estimating soil organic matter content from Hyperion reflectance images using PLSR, PCR, MinR and SWR models insemiarid regions of Iran[J]. Environmental Development, 2018, 25: 23-32.

[66] Casa R, Castaldi F, Pascucci S, et al. Potential of hyperspectral remote sensing for field scale soil mapping and precision agriculture applications[J]. Italian Journal of Agronomy, 2012, 7(4): e43.

[67] 廖钦洪, 顾晓鹤, 李存军, 等. 基于连续小波变换的潮土有机质含量高光谱估算[J]. 农业工程学报, 2012, 28(23): 132-139+298.

[68] 刘焕军, 张柏, 赵军, 等. 黑土有机质含量高光谱模型研究[J]. 土壤学报, 2007, (1): 27-32.

[69] 张霞, 孙友鑫, 尚坤, 等. 基于有机质特征谱段的土壤Cd含量高光谱遥感反演[J]. 农业机械学报, 2024, 55(1): 186-195.

[70] 邓昀, 王君, 陈守学, 等. 基于PCA-DBO-SVR的林地土壤有机质高光谱反演模型[J]. 光谱学与光谱分析, 2025, 45(2): 569-583.

[71] 周磊, 何秀芹, 贾德伟, 等. 基于CWT-CARS的黄土高原东部陕州区农田表层有机碳含量高光谱反演[J]. 遥感技术与应用, 2025, 40(1): 60-68.

[72] Padmavathi M, Manikandan M, Sumithra M G, et al. Impact of advanced sensing technologies in agriculture with soil, crop, climate and farmland-based approaches using internet of things[M]//Computational Intelligence in Internet of Agricultural Things. Cham: Springer Nature Switzerland, 2024: 109-146.

[73] 唐柜彪. 农业用地土壤重金属样本点数据精化方法研究[D]. 西安: 西安科技大学, 2021.

[74] 张东悦, 董士伟, 杨茂伟, 等. 场地土壤污染监测点空间优化布设方法[J]. 环境污染与防治, 2022, 44(8): 1074-1078+107.

[75] Hartono R, Yoeseph N M, Purnomo F A, et al. Portable internet of things-based soil nutrients monitoring for precision and efficient smart farming[J]. Bulletin of Electrical Engineering and Informatics, 2024, 13(5): 3326-3333.

[76] Canal Filho R, Molin J P. Spatial distribution as a key factor for evaluation of soil attributes prediction at field level using online near-infrared spectroscopy[J]. Frontiers in Soil Science, 2022, 2: 984963.

[77] 谢宝妮, 常庆瑞, 秦占飞. 县域土壤养分离群样点检测及其合理采样数研究[J]. 干旱地区农业研究, 2012, 30(2): 56-61.

[78] 马玉峰, 王文明, 贾宝全. 半干旱农牧交错区土壤养分空间变异特征分析——以内蒙古伊金霍洛旗为例[J]. 干旱区资源与环境, 2007, (11): 123-130.

[79] 周脚根, 赵春江, 王纪华, 等. 基于局部离群指数的土壤重金属污染评价方法[J]. 农业工程学报, 2010, 26(S1): 279-284.

[80] Shekhar S, Lu C T, Zhang P. Detecting graph-based spatial outliers: algorithms and applications (a summary of results)[C]//Proceedings of the seventh ACM SIGKDD international conference on Knowledge discovery and data mining. 2001: 371-376.

[81] Chawla S, Sun P. SLOM: a new measure for local spatial outliers[J]. Knowledge and Information Systems, 2006, 9: 412-429.

[82] 刘莘, 张绍良, 王飞, 等. 基于地统计学的空间离群点检测算法的研究[J]. 计算机应用研究, 2016, 33(12): 3700-3704.

[83] Hiremath A K, Kumar K R N, Malleshappa M H, et al. Assertion of soil data consistency by detecting and removing spatial outliers using iterative Techniques for precision agriculture[J]. IEEE Access, 2024.

[84] 张欢, 高小红. 复杂地形区土壤有机质空间变异性分析及制图[J]. 水土保持研究, 2020, 27(5): 93-100.

[85] Bhattacharjee S, Mitra P, Ghosh S K. Spatial interpolation to predict missing attributes in GIS using semantic kriging[J]. IEEE Transactions on Geoscience and Remote Sensing, 2013, 52(8): 4771-4780.

[86] Wang Y, Yang M, Wang M, et al. A novel potential outlier recognition approach considering local heterogeneity enhancement to improve the quality of soil datasets[J]. Geoderma, 2025, 454: 117200.

[87] 谢梦姣, 陈奇乐, 张俊梅, 等. 短距离样点对土壤呼吸空间变异预测精度的影响[J]. 中国生态农业学报, 2020, 28(3): 421-428.

[88] 潘瑜春, 刘巧芹, 陆洲, 等. 离群样点对土壤养分空间变异分析的影响研究[J]. 土壤学报, 2010, 47(4): 767-771.

[89] 樊桂菊, 李钊, 毛文华, 等. 基于工作空间的果园作业平台结构参数优化与试验[J]. 农业机械学报, 2021, 52(4): 34-42+265.

[90] 李憑峰, 谭煌, 王嘉航, 等. 滴灌水肥条件对樱桃产量、品质和土壤理化性质的影响[J]. 农业机械学报, 2017, 48(7): 236-246.

[91] Pan Y, Ren X, Gao B, et al. Global mean estimation using a self-organizing dualzoning method for preferential sampling[J]. Environmental Monitoring and Assessment, 2015, 187: 1-10.

[92] Ren X, Wei N, Gao B, et al. Experimental study on estimation of global mean with preferential spatial samples[J]. Intelligent Automation & Soft Computing, 2014, 20(4): 625-635.

[93] 李安娜, 马庆伟, 董士伟, 等. 兼顾面积属性与不确定性信息的样本点权重调整方法[J]. 农业机械学报, 2023, 54(05): 219-226.

[94] 张东悦, 朱庆伟, 董士伟, 等. 土壤Hg和NDVI全局均值估计的样本点权重调整方法[J]. 农业机械学报, 2022, 53(3): 338-345.

[95] 唐柜彪, 朱庆伟, 董士伟, 等. 农业用地土壤重金属样本点数据精化方法——以北京市顺义区为例[J]. 农业环境科学学报, 2020, 39(10): 2288-2296.