不同含水量尾砂的光谱特征与遥感模型

doi:10.3964/j.issn.1000-0593(2019)10-3096-06

Abstract
References
Related Citation (15)

Download: PDF (3486 KB)
Export: BibTeX | EndNote (RIS)

Abstract The tailing ponds are widely distributed in China. Once the surface tailings are under low moisture content, the tailing dust would cause severely environmental pollution driven by the wind action. Because of the large area and rapidly variation of moisture content of tailing ponds, the traditional method with the limitation on low efficiency, safety and high cost cannot meet the requirement of quick and dynamic moisture content monitoring. Currently, although the remote sensing technology based on spectral model can provide accurate prediction of soil moisture content, this model is not fit to tailing moisture content prediction because of the different characteristics and components between soil and tails. Therefore, the Fengshuigou tailing pond in Liaoning Province was selected as the study area. First, the samples of tailings at different moisture content were collected and configured. Then, the visible-near infrared spectra of the samples were measured and analyzed. Furthermore, the relationship between moisture contents and spectral characteristics was established. Finally, the remote sensing inversion model for moisture contents prediction was built and applied for mapping the moisture content in this study area. This study yielded the following results: （1） The moisture content has a significant effect on the spectral characteristics of tailings，and the reflectance decreased obviously as the moisture content increased. The longer the wavelength is, the more significant the effect of water content on the spectrum is. （2） The remote sensing model based on the spectral characteristic for tailings moisture content prediction was established. In terms of the band 6 and band 7 from Landsat8-OLI imagery, the ratio index (RTI), normalized difference index (NDTI) and difference index (DTI) of tailings were proposed and selected as the input data for the random forest model. By comparing the random forest model and Log reflectance model, the random forest model can generate more accurate predicting results. （3） The tailings moisture content map was generated by applying the random forest predicting model with spectral index based on the Landsat-OLI imagery. From the field verification, the coefficie of determination (R²), RMSE, RPD, and ARE is 0.798, 0.077, 1.970 and 20.1% respectively between the predicted and field measured moisture content. The results could provide an effective and real-time method for large scale moisture content predicting of the tailing ponds from the metamorphic iron ore area.

Key words：Visible and near-infrared spectrum; Tailings; Moisture content; Prediction

Received: 2018-09-10 Accepted: 2019-02-16

ZTFLH:

P237

Corresponding Authors: LIU Shan-jun E-mail: liusjdr@126.com

Cite this article:

URL:

https://www.gpxygpfx.com/EN/10.3964/j.issn.1000-0593(2019)10-3096-06 OR https://www.gpxygpfx.com/EN/Y2019/V39/I10/3096

[1] Liu R Z, Liu J, Zhang Z J, et al. International Journal of Environmental Research and Public Health, 2015, 12: 152.
[2] Lkhagvajargal B, Sonomdagva C, Byambatseren C, et al. Proceedings of the Mongolian Academy of Sciences, 2017, 57(4): 224.
[3] Hasan S, Montzka C, Rüdiger C, et al. ISPRS Journal of Photogrammetry and Remote Sensing, 2014, 91: 59.
[4] Łopatka A, Miturski T, Pudełko R, et al. Polish Journal of Soil Science, 2016, 49(1): 73.
[5] Parrens M, Wigneron J P, Richaume P, et al. Remote Sensing of Environment, 2016, 181: 122.
[6] Gao Z Q, Gao W, Chang N B. International Journal of Applied Earth Observation and Geo information，2011, 13(3): 495.
[7] WU Yong-feng, DONG Yi-wei, HU Xin, et al(武永峰, 董一威, 胡新，等). Spectroscopy and Spectral Analysis(光谱学与光谱分析), 2015, 35(12): 3416.
[8] Chang C W, Laird D, Hurburgh J, et al. Soil Science, 2005, 170: 244.
[9] Liu W, Baret F, Gu X, et al. International Journal of Remote Sensing, 2003, 24(10): 2069.
[10] Haubrock S N, Sabine C, Matthias K , et al. Journal of Applied Remote Sensing, 2008, 2: 23.
[11] Haubrock S N, Chabrillat S, Lemmnitz C, et al. International Journal of Remote Sensing, 2010, 29: 13.
[12] Zeng W, Xu C, Huang J H, et al. Soil Physics & Hydrology, 2016, 6: 1496.
[13] Yuanda Z, David C W, Somsubhra C, et al. Journal of Hydrology, 2010, 391: 133.
[14] Leila H, Alfonso T, Austin J, et al. Remote Sensing, 2015, 7: 2627.
[15] Daniel R G, Wade A W, Andrew J F, et al. Journal of Terramechanics, 2016, 65: 85.
[16] Wang X P, Zhang F, Ding J L, et al. Science of the Total Environment, 2018, 615: 918.
[17] Roberts J, Cozzolino D. Trends in Analytical Chemistry, 2016, 83: 25.
[18] Ouerghemmi W, Gomez C, Naceur S. Remote Sensing of Environment, 2016, 181: 251.
[19] Peng J, LI X, Zhou Q. Journal of Remote Sensing, 2013, 6: 1396.
[20] Shuho N, Yasushi Y. Ore Geology Reviews, 2017, 83: 312.