Plant litter estimation and its correlation with sediment concentration in the Loess Plateau
Author and article information
Abstract
Background. Sediment concentration in the water of the loess Plateau region has dramatically decreased during the past two decades. Plant litter is considered to be one of the most important factors for this change. Existing remote sensing studies that focus on plant litter mainly use extraction methods based on vegetation indices or changes in the plant litter. Few studies have conducted time series analyses of plant litter or considered the correlation between plant litter and soil erosion. In addition, social factors are not given enough consideration in the remote sensing and soil community. Methods. This study performs time series estimation of plant litter by integrating three-scale remotely sensed data and a random forest (RF) modeling algorithm. Predictive models are used to estimate the spatially explicit plant litter cover for the entire Loess Plateau over the last two decades (2000–2018). Then, the sediment concentration in the water was classified into 9 grades based on environmental and social-economic factors. Results. Our results demonstrate the effectiveness of the proposed predictive models at the regional scale. The areas with increased plant litter cover accounted for 67% of the total area, while the areas with decreased plant litter cover accounted for 33% of the total area. In addition, plant litter is demonstrated to be one of the top three factors contributing to the decrease in the river sediment concentration. Social-economic factors were also important for the decrease of the sediment concentration in the water, for example, the population of the rural area.
Cite this as
2019. Plant litter estimation and its correlation with sediment concentration in the Loess Plateau. PeerJ Preprints 7:e27891v1 https://doi.org/10.7287/peerj.preprints.27891v1Author comment
This is a submission to PeerJ for review.
Sections
Additional Information
Competing Interests
The authors declare that they have no competing interests.
Author Contributions
Qian Li analyzed the data, prepared figures and/or tables.
Ligang Ma conceived and designed the experiments, performed the experiments, authored or reviewed drafts of the paper, approved the final draft.
Suhong Liu performed the experiments.
Adilai Wufu analyzed the data, contributed reagents/materials/analysis tools.
Yinbo Li analyzed the data, contributed reagents/materials/analysis tools.
Shengtian Yang performed the experiments.
Xiaodong Yang the teacher Xiaodong Yang made suggestions for the revision of the paper.
Data Deposition
The following information was supplied regarding data availability:
FigShare
DOI:10.6084/m9.figshare.9415880
https://figshare.com/articles/Plant_litter_cover/9415880/2
The 01_The_First_Upscaling_RF.R implements the first-level random forest modeling. The input data is 01_The_First_Upscaling folder, where 01_ModelData folder is modeling data and 02_PredictData folder is data used for prediction.
The 02_The_Second_Upscaling_RF.R implements the second-level random forest modeling. The input data is 02_The_Second_Upscaling folder, where 01_ModelData folder is modeling data and 02_PredictData folder is data used for prediction.
The 03_Slope_and_F.R script is to calculate the annual average trend of change and F statistics of the plant litter cover of 2000-2018.
The 04_Classification_C50.R realizes the classification of the sediment concentration grade and calculates the KAPPA coefficient. The file 04_C50.xlsx is the input data.
Data_Processing_Introductions.docx explains some of the considerations for running R scripts.
Funding
This work was supported by the National Key Research and Development Program of China (2016YFC0402409) and the National Natural Science Foundation of China (41661079, U1603241, 41771470). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
