A satellite-based hybrid model for trophic state evaluation in inland waters across China.

Eutrophication is one of the major threats to the inland water ecosystem. Satellite remote sensing provides a promising way to monitor trophic state at large spatial scale in an efficient manner. Currently, most satellite-based trophic state evaluation approaches have focused on water quality parameters retrieval (e.g., transparency, chlorophyll-a), based on which trophic state was evaluated. However, the retrieval accuracies of individual parameter do not meet the demand for accurate trophic state evaluation, especially for the turbid inland waters. In this study, we proposed a novel hybrid model to estimate trophic state index (TSI) by integrating multiple spectral indices associated with different eutrophication level based on Sentinel-2 imagery. The TSI estimated by the proposed method agreed well with the in-situ TSI observations, with root mean square error (RMSE) of 6.93 and mean absolute percentage error (MAPE) of 13.77%. Compared with the independent observations from Ministry of Ecology and Environment, the estimated monthly TSI also showed good consistency (RMSE=5.91,MAPE=10.66%). Furthermore, the congruent performance of the proposed method in the 11 sample lakes (RMSE=5.91,MAPE=10.66%) and the 51 ungauged lakes (RMSE=7.16,MAPE=11.56%) indicated the favorable model generalization. The proposed method was then applied to assess the trophic state of 352 permanent lakes and reservoirs across China during the summers of 2016-2021. It showed that 10%, 60%, 28%, and 2% of the lakes/reservoirs are in oligotrophic, mesotrophic, light eutrophic, and middle eutrophic states respectively. Eutrophic waters are concentrated in the Middle-and-Lower Yangtze Plain, the Northeast Plain, and the Yunnan-Guizhou Plateau. Overall, this study improved the trophic state representativeness and revealed trophic state spatial distribution of Chinese inland waters, which has the significant meanings for aquatic environment protection and water resource management.

Estimation of the soil arsenic concentration using a geographically weighted XGBoost model based on hyperspectral data.

Considering the high toxicity of arsenic (As), its contamination of soil represents an alarming environmental and public health issue. Existing soil heavy metal concentration estimation models based on hyperspectral data ignore the spatial nonstationarity of the relationship between the soil spectrum and heavy metal concentration. A novel model (geographically weighted eXtreme gradient boosting or GW-XGBoost model) combining geographically weighted regression (GWR) method with XGBoost algorithm was proposed. The northeast district of Beijing, China, was chosen as a case study area to assess the effectiveness of the proposed model. The GW-XGBoost model was established to estimate the As concentration based on the typical spectrum of As and the spatial correlation between the spectrum and As concentration obtained using the GWR method, and the result was compared to that obtained with the XGBoost and GWR models. The accuracy of the GW-XGBoost model was obviously better than that of the other models (R2GW-XGBoost = 0.90, R2XGBoost = 0.48, and R2GWR = 0.74). Therefore, the proposed model is reliable, as it considers the spatial correlation between the spectrum and As concentration.

Spatiotemporal dynamics of suspended particulate matter in the Bohai Sea, China over the past decade from the space perspective.

Suspended particulate matter (SPM) concentration is an important biogeochemical parameter for water quality assessment and morphodynamic studies. In this study, the four recent SPM retrieval models developed for Bohai Sea were evaluated using in situ datasets, and the best performing model was selected to investigate the spatiotemporal dynamics of SPM in Bohai Sea from 2011 to 2021 based on 1164 satellite imageries. The results indicated that the satellite-derived SPM concentrations had a high accuracy (R2 = 0.86, relative percentage difference = 33.71 %). The SPM concentrations in the Bohai Sea demonstrated a significant decadal decreasing trend (0.503 mg/L/yr), and the distribution area with low SPM (<30 mg/L) increased by 3.29 % annually. The southern Bohai Sea declined observably, involving the Bohai Bay (2.07 mg/L/yr), Laizhou Bay (1.916 mg/L/yr), and central Bohai Sea (-0.661 mg/L/yr). Monthly SPM was characterized by significant seasonality. The SPM circulation pattern in the Bohai Strait was generally northerly inflow and southerly outflow. Significant wave heights (Hs) dominated the SPM variations and explained 58.9 % of monthly SPM changes in the Bohai Sea. The strong waves reduction was the main reason for the decadal decline of SPM concentrations. Wind waves associated with monsoons controlled seasonal variations of SPM and promoted the output in winter through the southern Bohai Strait. Storms could cause a sharp increase in SPM concentrations, especially in Bohai Bay and Laizhou Bay which were highly sensitive to northerly winds and strong waves. After the storm ended, the effects of short-duration storm might fade away within a few hours, while that of long-duration storm could last for 2-3 days. High sediment transport from Yellow River (>500 × 104 t/M) controlled 74.8 % of monthly SPM variations within 3-km area off the estuary, 45 % of that within 5-km area, and 28.4 % of that within 10-km area.

The impact assessment of hydro-biological connectivity changes on the estuary wetland through the ecological restoration project in the Yellow River Delta, China.

Yellow River Delta (YRD) is one of the youngest delta with complex hydrological and biological connectivity in the world, where offers habitats to the famous waterfowls in the Eastern Asia. Meanwhile, one specific ecological restoration project named as the "Wuwanmu" and followed by the "Shiwanmu" within the National Nature Reserve of the Yellow River Delta (NNRYRD) complicated the hydrological and biological connectivity. How to quantitatively evaluate the extent of coastal wetland affected by the project will be a difficult problem. Hence the authors presented three innovative models of the Marine Connectivity Change Index (MCCI), the Coupling Index of Hydro-biological Connectivity (CIHBC), and the Assessment Index of Suitability on Bird Habitats (AISBH). After the project, the habitat of Phragmites australis has been restored effectively with the increased area of 24.59%, while the habitat of Suaeda salsa as the native species lost largely with decreased area of 84.62%. And the tidal channel having been cut off by the project resulted in isolating the buildup restoration area from seawater, and reshaping completely the plant habitat environment. So the hydrological and biological connectivity has been changed largely with the 47.79% decreased MCCI area and the 16.3% decreased zero-valued CIHBC area. However the AISBH non-zero-valued area increased 10.7%, and with the hidden worry of the decreased Grallatores number. From the connectivity prospective, three models presented a significant methodology to evaluate the complex impact on the estuary wetland habitat caused by the restoration project. In the long run, the ecological impacts should be highlighted to the change of tidal channel and the corresponding tidal issues, and the continuous and big loss of native plant spices such as S. salsa. The further study needs to explore the longer-term assessment of the ecological restoration project and its multiple effect in the future.

Human impact on suspended particulate matter in the Yellow River Estuary, China: Evidence from remote sensing data fusion using an improved spatiotemporal fusion method.

Transport of suspended particulate matter (SPM) in estuarine waters plays an important role in regulating erosion-accretion and biogeochemical processes. In the Yellow River Estuary (YRE), artificial water and sediment regulation scheme (WSRS) and coastal engineering structures are the 2 typical anthropogenic activities affecting the spatiotemporal dynamics of estuarine SPM. The monitoring of SPM transport affected by such human activities requires SPM mapping at both high spatial and high temporal resolutions. In this study, we presented an improved Flexible Spatiotemporal Data Fusion (FSDAF) strategy with consideration of highly dynamic SPM variations in estuarine waters, and generated 30-m hourly SPM concentrations based on Landsat 8 OLI and GOCI datasets. The new strategy produced higher SPM estimation accuracy than the original FSDAF, with the relative percentage difference (RPD) decreasing from 29.75% to 5.31% using GOCI-derived hourly SPM as reference. With in situ SPM measurements as reference, the fused SPM concentrations had an RMSE of 12.09 mg/L and an RPD of 27.17%. Investigation of interday SPM variations before, during, and after the WSRS in 2018 revealed that the first WSRS significantly increased the SPM concentration and plume extent; new wetland with an area of 12.56 km2 was formed due to sediment accretion near the river mouth. The two groins offshore from the coastlines on the north and south sides of YRE exhibited obvious sediment trapping effects in that higher SPM concentrations on one side of each groin were found regardless of the turbidity modes and diurnal SPM variations; the trapping effects were associated with the number of groins and groin length. Intraday variations of SPM were influenced by tidal currents, with plume direction following the ebb and flooding tidal current direction. The inter- and intraday characteristics of the 30-m hourly SPM dynamics facilitate the detailed analysis of the sediment transport associated with human activities.

Spatiotemporal dynamics of suspended particulate matter in the Yellow River Estuary, China during the past two decades based on time-series Landsat and Sentinel-2 data.

Twenty-two years of suspended particulate matter (SPM) concentrations in the Yellow River estuary and adjacent sea, China were derived from 532 Landsat and Sentinel 2A/B satellite images. Optimal SPM retrieval model was selected by comparing five state-of-art models using 79 in-situ datasets and recalibrated to ensure consistency among multiple-sensor-derived SPM concentrations. SPM in the estuary, in South Bohai Bay, and Laizhou Bay exhibited distinct temporal variations. 73% and 52% of the interannual and monthly SPM variations near the river mouth were explained by riverine water and sediment discharge, showing impact of the operation of the Xiaolangdi Reservoir and Water-Sediment Regulation Scheme. Land area accretion and erosion in river delta are associated with SPM variation. Riverine impacts on SPM rapidly declined off-shore because of the rapid deposition of the coarse-grain sediment. Ocean current and wind-wave forces explained high concentrations and intra-annual variations of SPM in the South Bohai Bay and Laizhou Bay.

Urban Land Use and Land Cover Classification Using Novel Deep Learning Models Based on High Spatial Resolution Satellite Imagery.

Urban land cover and land use mapping plays an important role in urban planning and management. In this paper, novel multi-scale deep learning models, namely ASPP-Unet and ResASPP-Unet are proposed for urban land cover classification based on very high resolution (VHR) satellite imagery. The proposed ASPP-Unet model consists of a contracting path which extracts the high-level features, and an expansive path, which up-samples the features to create a high-resolution output. The atrous spatial pyramid pooling (ASPP) technique is utilized in the bottom layer in order to incorporate multi-scale deep features into a discriminative feature. The ResASPP-Unet model further improves the architecture by replacing each layer with residual unit. The models were trained and tested based on WorldView-2 (WV2) and WorldView-3 (WV3) imageries over the city of Beijing. Model parameters including layer depth and the number of initial feature maps (IFMs) as well as the input image bands were evaluated in terms of their impact on the model performances. It is shown that the ResASPP-Unet model with 11 layers and 64 IFMs based on 8-band WV2 imagery produced the highest classification accuracy (87.1% for WV2 imagery and 84.0% for WV3 imagery). The ASPP-Unet model with the same parameter setting produced slightly lower accuracy, with overall accuracy of 85.2% for WV2 imagery and 83.2% for WV3 imagery. Overall, the proposed models outperformed the state-of-the-art models, e.g., U-Net, convolutional neural network (CNN) and Support Vector Machine (SVM) model over both WV2 and WV3 images, and yielded robust and efficient urban land cover classification results.