Estimation of total iron content in floodplain soils using VNIR spectroscopy – a case study in the Le'an River floodplain, China

Abstract

Accumulation of heavy metals has become a serious environmental issue in China, especially in the floodplains downstream from mining and smelting sites. Estimation of total iron (Fe) content at the regional scale becomes particularly important because of the heavy metal sorption of Fe oxide. A novel method for estimating total Fe content is proposed using visible and near-infrared (VNIR) spectroscopy and partial least squares regression (PLSR). Our study focuses on the Le'an River floodplain, Jiangxi Province, China, which houses the largest copper mining corporation in China, as this area has suffered a series of environmental setbacks because of the extraction of non-ferrous metals. Our study employs PLSR to summarize the relationship between VNIR reflectance spectra and the total Fe contents of collected soil samples. More specifically, our study aims to (1) explore the correlation between VNIR and total Fe content, (2) assess the relationship between VNIR determination of total Fe content and the preprocessing of soil samples and (3) evaluate the performance of data transformation methods in PLSR. The PLSR model with transformed total Fe content and continuum removal spectra was finally chosen for estimating the total Fe content from both pretreated soil samples (coefficient of determination for prediction, = 0.66) and soil samples without pretreatment ( = 0.55). Therefore, VNIR spectroscopy could be an alternative method for estimating total Fe content at the regional scale.