摘要
一位新闻记者-机器人与机器学习的工作人员新闻编辑每日新闻-机器学习的新研究是一篇报告的主题。据中国人民共和国北京的新闻报道,NewsRx记者研究表明:“土壤有机碳(SOC)对全球碳循环和环境可持续发展至关重要。同时,快速、方便的遥感技术已成为监测土壤有机碳含量的重要手段之一。”新闻记者从中国农业大学的研究中得到一句话：“现在，”在基于稀缺地面点的高精度复杂空间关系反演SOC含量方面存在局限性，受同一物质的不同比值以及地形和环境（如植被和气候）的影响，SOC含量与遥感光谱关系的空间差异的制约。利用两点机器学习(TPML)m方法,结合Sentinel-1、Sentinel-2、地形和环境的衍生变量,对黑龙江省海伦县的SOC含量进行反演,结果表明,TPML方法反演精度最高。其次是随机森林法、梯度boosting回归树法、偏最小二乘回归法、支持向量机法,TPML法的平均误差R、平均误差MAE、平均误差RMSE和平均误差RPD分别为0.854%、0.384%、0.558%和1.918.。通过对反演结果的实际误差和理论误差的比较,发现TPML反演分辨率为10m的SOC含量比其他模型更平滑、更真实,与不同土地利用类型的SOC含量分布一致。
Abstract
By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News-New research on Machine Learning is th e subject of a report. According to news reporting from Beijing, People's Republ ic of China, by NewsRx journalists, research stated, "Soil organic carbon (SOC) is vital for the global carbon cycle and environmentally sustainable development . Meanwhile, the fast, convenient remote sensing technology has become one of th e notable means to monitor SOC content." The news correspondents obtained a quote from the research from China Agricultur al University, "Nowadays, limitations are found in the inversion of SOC content with high-precision and complex spatial relationships based on scarce ground sam ple points. It is restrained by the spatial difference in the relationship betwe en SOC content and remote sensing spectra due to the problem of different spectr a for the same substance and the influence of topographic and environment (e.g. vegetation and climate). In this regard, the two-point machine learning (TPML) m ethod, which can overcome above problems and deal with complex spatial heterogen eity of relationships between SOC and remote sensing spectra, was used to invert the SOC content in Hailun County, Heilongjiang Province, combined with derived variables from Sentinel-1, Sentinel-2, topography and environment. Based on 10-f old cross-validation and t-test, results indicated that the TPML method boasts t he highest inversion accuracy, followed by random forest, gradient boosting regr ession tree, partial least squares regression and support vector machine. The av erage r, MAE, RMSE, and RPD of TPML were 0.854, 0.384 %, 0.558 % , and 1.918. Further, the TPML method has been proven to be equal to evaluating the uncertainty of inversion results, by comparing the actual and theoretical er ror of the inversion result in one subset. The spatial inversion result of SOC c ontent with 10 m resolution by TPML is smoother and has more real details than o ther models, which are consistent with the distribution of SOC content in differ ent land use types."
NETL
NSTL