Chelant-enhanced phytoextraction method has been put forward as an effective soil remediation method, whereas the heavy metal leaching could not be ignored. In this study, a cropping-leaching experiment, using soil columns, was applied to study the metal leaching variations during assisted phytoextraction of Cd- and Pb-polluted soils, using seedlings of Zea mays, applying three different chelators (EDTA, EDDS, and rhamnolipid), and artificial rainfall (acid rainfall or normal rainfall). It showed that artificial rainfall, especially artificial acid rain, after chelator application led to the increase of heavy metals in the leaching solution. EDTA increased both Cd and Pb concentrations in the leaching solution, obviously, whereas EDDS and rhamnolipid increased Cd concentration but not Pb. The amount of Cd and Pb decreased as the leaching solution increased, the patterns as well matched LRMs (linear regression models), with R-square (R2) higher than 90 and 82% for Cd and Pb, respectively. The maximum cumulative Cd and Pb in the leaching solutions were 18.44 and 16.68%, respectively, which was amended by EDTA and acid rainwater (pH 4.5), and followed by EDDS (pH 4.5), EDDS (pH 6.5), rhamnolipid (0.5 g kg−1 soil, pH 4.5), and rhamnolipid (pH 6.5).
This study aims to apply Moderate Resolution Imaging Spectroradiometer (MODIS Data) to monitor water quality parameters including chlorophyll-a, secchi disk depth, total phosphorus and total nitrogen at Chaohu Lake. In this paper, multivariate regression analysis, Back Propagation neural networks (BPs), Radial Basis Function neural networks (RBFs) and Genetic Algorithms-Back Propagation (GA-BP) were applied to investigate the relationships between water quality parameters and the MODIS bands combinations. The study results indicated that a simple, efficient and acceptable model could be established through multivariate regression analysis, but the model precision was relatively low. In comparison, BPs, RBFs and GA-BP were significantly advantageous in terms of sufficient utilization of spectra information and model reliance. The relative errors of BPs, RBFs and GA-BP were below 35%. Based on method comparison, it can be concluded that GA-BP is more suitable for simulation and prediction of water quality parameters by applying genetic algorithm to optimize the weight value of BP network. This study demonstrates that MODIS data can be applied for monitoring some of the water quality parameters of large inland lakes. 相似文献