Estimation of Suspended Sediment Concentration in the Yangtze Main Stream Based on Sentinel-2 MSI Data

Zhang, Chenlu; Liu, Yongxin; Chen, Xiuwan; Gao, Yu

doi:10.3390/rs14184446

Cited by 20 publications

(23 citation statements)

References 69 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The multiple linear regression method estimates forest biomass by using the extracted important features as independent variables and establishing a correlation with the measured AGB in the sample plots. This method is simple and intuitive and can quantitatively describe the linear relationships between variables (Fang et al, 2019; Su et al, 2017; Zhang et al, 2018).…”

Section: Methodsmentioning

confidence: 99%

“…As what we have described in Section 2.3.1, random forest is a classification regression method proposed by Breiman to build Bagging parallelized integrated learning with decision tree learners (Breiman, 2001; Fang et al, 2019). The method randomly takes out a sample in the given data set and puts it into the training set by the self‐service sampling method, then puts the sample into the initial data set so that it is likely to be randomly sampled next time.…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Estimation of forest above‐ground biomass in Guangxi, China, by integrating forest age and stack learning

Liu

Yue

et al. 2023

Land Degrad Dev

View full text Add to dashboard Cite

With the implementation of large‐scale ecological restoration projects, Southwest China has become one of the fastest forest growth areas in the world in terms of vegetation cover and above‐ground biomass (AGB). It is expected to be a potential area for achieving the carbon neutrality target in China. Accurate estimation of forest AGB is becoming an increasingly urgent necessity for carbon neutrality and forest management. However, due to the complex geological background, there is significant uncertainty in estimating forest AGB in the southwestern region, which generally results in underestimating carbon sinks from forest restoration. To address the issue, we propose a method by incorporating forest age information and stack learning technique to estimate forest AGB. Based on remote sensing, forest inventory and in situ forest biomass data, three fundamental methods (Multiple regression, Random forest and Support vector machine) are employed and compared to build the AGB estimation model with vegetation indices, texture feature factors and forest age. Optimal basic models are further enhanced by integration learning to improve the estimation performance and then applied to the study area Guangxi to obtain regional AGB information of different forest types. The results show that: (1) forest age plays a vital role in reducing the uncertainty of AGB estimation. By incorporating forest age information, R2 of AGB estimation is improved by 0.07–0.27 and RMSE is decreased by 16.35%–47.47% for different forest types; (2) with R2 value >0.78, random forest model outperforms support vector machine and multiple linear regression models. Compared with the single optimal model, integration model by stack learning further enhances R2 of estimation by 0.02–0.03 and decreases RMSE by 5.20%–14.89%. (3) The total forest AGB in Guangxi is 988.17 Tg and the average forest AGB level is 73.30 t/ha. Natural broadleaf forest has the highest AGB level (86.75 t/ha), followed by natural coniferous forest (81.19 t/ha), planted coniferous forest (63.23 t/ha) and planted eucalyptus forest (49.71 t/ha). AGB level in karst areas is lower than that in non‐karst areas due to soil and water constraints. The majority of plantation forests in Guangxi is in the early and middle stages of forest succession, with the rapid growth of forest AGB, and has hence significant potential as carbon sinks. This study indicates that stack learning and incorporation of forest age could significantly decrease the uncertainty of forest AGB estimation. Our study helps to provide more accurate AGB information for karst ecological project management and regional carbon neutrality assessment.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Estimation of forest above‐ground biomass in Guangxi, China, by integrating forest age and stack learning

Liu

Yue

et al. 2023

Land Degrad Dev

View full text Add to dashboard Cite

show abstract

“…To eliminate the influence of non‐water pixels such as thin clouds and boats at the hydrological monitoring sites, surface reflectance data were extracted for 3 × 3 pixels surrounding the hydrological stations along the river channels. Due to the high reflectance values of water bodies in the red band and low reflectance values in the near‐infrared band, the Normalized Difference Vegetation Index (NDVI) has been proven useful for identifying pure water bodies (Fang et al., 2019). The minimal NDVI value in the 3 × 3 pixels near the hydrographic station was used to pinpoint pure water bodies and accordingly their surface reflectance values were extracted.…”

Section: Methodsmentioning

confidence: 99%

Google Earth Engine‐Based Estimation of the Spatio‐Temporal Distribution of Suspended Sediment Concentrations in a Multi‐Channel River System of the Yangtze River Basin

Zhao,

Huang,

et al. 2023

Water Resources Research

View full text Add to dashboard Cite

Lowland multi‐channel alluvial river systems are highly variable in frequency and magnitude of floodplain inundation and are vulnerable to human activities such as damming. In the Yangtze River Basin, the Three Gorges Dam (TGD) has trapped >80% of upstream sediment supply, causing downstream scouring and rapid geomorphic changes in river and its floodplain lakes. Suspended sediment concentration (SSC) is widely used to monitor river morphodynamics, but traditional measurements of SSC are time consuming, costly and difficult to quantify SSC in a large spatial scale. Using Google Earth Engine and in situ observed hydrological data, we created a multiple linear regression model to map SSC in the multi‐channel system Songzi River of the Yangtze. The new SSC predictive model achieved high accuracy (R2: 0.87) and showed opposite downstream trends in SSC during peak flood years and normal flood years. For the first time, we found that in the peak flood year of 1998 the study rivers exhibit a downstream increase trend of SSC with an abrupt increase in their middle reach, while SSC in normal flood years experiences a downstream decline with minimal changes. A prominent difference in SSC is also revealed after the TGD with a reduction of 60%. Furthermore, SSC in the closely hydrologically connected lakes is more dynamic than the less connected lakes. Our study demonstrates that the proposed method enables to rapidly quantify the spatio‐temporal SSC distribution in the multi‐channel systems on the floodplain of large alluvial rivers and highlights the importance of connectivity in regulating SSC dynamics.

show abstract

“…With better spatial resolution and development of assessment techniques, scope of application can now be extended to smaller reservoirs and longer river reaches. Zhang et al (2022) used the multispectral imagery and two-stage non-linear relationship between suspended sediment concentration and reflectance in various bands, valid for the range 2-850 PPM, and applied it to the whole mainstream Yangtze River.…”

Section: Earth Observation Assisted Monitoringmentioning

confidence: 99%

Monitoring and Modelling of Sediment Flushing : A Review

Petkovšek

2023

SRC

View full text Add to dashboard Cite

With ever decreasing potential for suitable new dam sites, sustainable use of existing water reservoirs is of paramount importance. In absence of appropriate measures, reservoir storage is continually reduced due to sedimentation. One option to remove sediment deposits is hydraulic flushing. During the flushing operation, bottom outlets are open and water and sediment released. Whether flushing successfully removes sediment depends on a number of factors, such as bottom outlets capacity, reservoir shape and water availability. Modelling is often used to assess viability of flushing for sediment management in the reservoir, as well as to design the operations and optimize their scheduling. One-dimensional numerical models are preferred for long term simulations, assessments on of a large number of scenarios, and optimization studies. Two- and three-dimensional numerical models and physical models can be used, each on their own or in combination as hybrid models, to understand local scouring near the gates and other details of operation. Monitoring of flushing operations can help improving their efficiency while at the same time limit downstream impacts. General monitoring of the reservoir and its catchment can help understanding the sedimentation problem and thus facilitate preparation of efficient sediment management strategies. Live monitoring of sediment concentrations is possible with modern equipment though not without challenges, and reservoir survey can be performed faster. Earth observation techniques are also an attractive option, allowing to monitor large areas and areas of difficult access, as well as to provide historical information going back several decades. This paper reviews monitoring and modelling approaches published in the literature, as well as presents some previously unpublished analyses.

show abstract

Estimation of Suspended Sediment Concentration in the Yangtze Main Stream Based on Sentinel-2 MSI Data

Cited by 20 publications

References 69 publications

Estimation of forest above‐ground biomass in Guangxi, China, by integrating forest age and stack learning

Estimation of forest above‐ground biomass in Guangxi, China, by integrating forest age and stack learning

Google Earth Engine‐Based Estimation of the Spatio‐Temporal Distribution of Suspended Sediment Concentrations in a Multi‐Channel River System of the Yangtze River Basin

Monitoring and Modelling of Sediment Flushing : A Review

Contact Info

Product

Resources

About