(1. 中南大學(xué) 有色金屬成礦預(yù)測(cè)與地質(zhì)環(huán)境監(jiān)測(cè)教育部重點(diǎn)實(shí)驗(yàn)室，長(zhǎng)沙 410083；
2. 中南大學(xué) 有色資源與地質(zhì)災(zāi)害探查湖南省重點(diǎn)實(shí)驗(yàn)室，長(zhǎng)沙 410083；
3. 中南大學(xué) 地球科學(xué)與信息物理學(xué)院，長(zhǎng)沙 410083)
摘 要: 與傳統(tǒng)的土壤重金屬含量定量研究方法相比，高光譜遙感技術(shù)具有成本低、效率高、探測(cè)范圍廣、宏觀性強(qiáng)、可動(dòng)態(tài)監(jiān)測(cè)等優(yōu)勢(shì)。基于空-天-地研究視角，分析了土壤重金屬高光譜數(shù)據(jù)特征、預(yù)處理方法與技術(shù)流程、應(yīng)用條件與范圍，論述了基于高光譜遙感技術(shù)開展土壤重金屬監(jiān)測(cè)的發(fā)展歷程。通過歸納基于土壤重金屬、土壤活性成分的光譜特征反演，發(fā)現(xiàn)土壤有機(jī)質(zhì)、鐵氧化物、黏土礦物等含量是導(dǎo)致土壤特征波段差異的關(guān)鍵因素。總結(jié)了土壤中常見活性成分及重金屬的特征波譜，認(rèn)為350~2500 nm是預(yù)測(cè)土壤成分含量的主要特征波段范圍。影響反演精度的關(guān)鍵因素包括土壤高光譜響應(yīng)機(jī)制、高光譜數(shù)據(jù)質(zhì)量、重金屬賦存狀態(tài)以及反演建模方法，改進(jìn)光譜優(yōu)化方法、構(gòu)建高效的反演模型、明確重金屬光譜特征及其賦存機(jī)理是進(jìn)一步提升土壤重金屬高光譜反演精度的有效途徑和方式。為適應(yīng)土壤重金屬含量定量監(jiān)測(cè)研究需要，未來高光譜遙感技術(shù)將會(huì)朝著定量化、宏觀化、主動(dòng)化、現(xiàn)場(chǎng)化方向發(fā)展。另外，數(shù)據(jù)多源化、非線性方法與線性方法融合、顧及重金屬和活性成分的多特征波段，也是今后高光譜遙感技術(shù)發(fā)展的重要趨勢(shì)。
關(guān)鍵字: 土壤；重金屬；活性成分；高光譜遙感；定量反演；建模方法
（1. Key Laboratory of Metallogenic Prediction of Nonferrous Metals and Geological Environment Monitoring, Ministry of Education, Central South University, Changsha 410083, China;
2. Hunan Key Laboratory of Nonferrous Resources and Geological Hazards Exploration, Central South University, Changsha 410083, China;
3. School of Geosciences and Info-Physics, Central South University, Changsha 410083, China）
Abstract:Compared with the traditional methods for quantitative study of soil heavy metal content, the hyperspectral remote sensing technology has greater advantages, such as lower cost, higher efficiency, wider detection range, strong macroscopic property, and dynamic monitoring, etc. Based on air-sky-earth research perspective, this paper analyzed the characteristics, pre-processing methods and technical processes, application conditions and scope of soil heavy metal hyperspectral data; it discussed the development of soil heavy metal monitoring based on hyperspectral remote sensing technology. By analyzing the inversion of spectral features based on soil heavy metals and soil active ingredients, we found that the contents of organic matter, iron oxides, and clay minerals in soil are the key factors leading to the differences in soil characteristic wavebands. In this paper, by summarizing the characteristic wave spectra of common active components and heavy metals in soil, it shows that 350-2500 nm is the main waveband range for predicting their contents, and the characteristic wavebands are very susceptible to soil types. The key factors affecting the inversion accuracy involve the soil hyperspectral response mechanism, hyperspectral data quality, occurrence state of heavy metals, and inversion modeling methods. The effective ways to further improve the accuracy of hyperspectral inversion of soil heavy metals include improving the spectral optimization method, constructing an efficient inversion model, and clarifying the spectral characteristics and occurrence mechanism of heavy metals. The future development of hyperspectral remote sensing technology will be characterized by quantitative, active, macroscopic and on-site. In addition, the data multi-source, fusion of non-linear and linear methods, and considering of multi-featured waveband information are also important trends for hyperspectral remote sensing technology in the future.
Key words: soil; heavy metals; active components; hyperspectral remote sensing; quantitative inversion; modeling method
