High-spatial-resolution images play an important role in land cover classification, and object-based image analysis (OBIA) presents a good method of processing high-spatial-resolution images. Segmentation, as the most important premise of OBIA, significantly affects the image classification and target recognition results. However, scale selection for image segmentation is difficult and complicated for OBIA. The main challenge in image segmentation is the selection of the optimal segmentation parameters and an algorithm that can effectively extract the image information. This paper presents an approach that can effectively select an optimal segmentation scale based on land object average areas. First, 20 different segmentation scales were used for image segmentation. Next, the classification and regression tree model (CART) was used for image classification based on 20 different segmentation results, where four types of features were calculated and used, including image spectral bands value, texture value, vegetation indices, and spatial feature indices, respectively. WorldView-3 images were used as the experimental data to verify the validity of the proposed method for the selection of the optimal segmentation scale parameter. In order to decide the effect of the segmentation scale on the object area level, the average areas of different land objects were estimated based on the classification results. Experiments based on the multi-scale segmentation scale testify to the validity of the land object’s average area-based method for the selection of optimal segmentation scale parameters. The study results indicated that segmentation scales are strongly correlated with an object’s average area, and thus, the optimal segmentation scale of every land object can be obtained. In this regard, we conclude that the area-based segmentation scale selection method is suitable to determine optimal segmentation parameters for different land objects. We hope the segmentation scale selection method used in this study can be further extended and used for different image segmentation algorithms.

Research on the consistency of suspended particulate matter (SPM) concentration retrieved from multisource satellite sensors can serve as long-time monitoring of water quality. To explore the influence of the atmospheric correction (AC) algorithm and the retrieval model on the consistency of the SPM concentration values, Landsat 8 Operational Land Imager (OLI) and Sentinel 2 MultiSpectral Imager (MSI) images acquired on the same day are used to compare the remote sensing reflectance (Rrs) SPM retrieval values in two high-turbidity lakes. An SPM retrieval model for Shengjin Lake is established based on field measurements and applied to OLI and MSI images: two SPM concentration products are highly consistent (R² = 0.93, Root Mean Squared Error (RMSE) = 20.67 mg/L, Mean Absolute Percentage Error (MAPE) = 6.59%), and the desired results are also obtained in Chaohu Lake. Among the four AC algorithms (Management Unit of the North Seas Mathematical Models (MUMM), Atmospheric Correction for OLI’lite’(ACOLITE), Second Simulation of Satellite Signal in the Solar Spectrum (6S), Landsat 8 Surface Reflectance Code & Sen2cor (LaSRC & Sen2cor)), the two Rrs products, as well as the final SPM concentration products retrieved from OLI and MSI images, have the best consistency when using the MUMM algorithm in SeaWIFS Data Analyst System (SeaDAS) software. The consistency of SPM concentration values retrieved from OLI and MSI images using the same model or same form of models is significantly better than that retrieved by applying the optimal models with different forms.

Landscape simulation and prediction are crucial for understanding the dynamic evolution and future trends of wetlands. However, only a few existing studies have focused on the applicability and limitations of commonly used land-use/cover change (LUCC) simulation models in lake wetland landscapes. Taking Shengjin Lake Reserve in China as the study area, we firstly analyzed landscape variations during 2010–2020 using multisource remote sensing images. Then, the patch-generating land-use simulation (PLUS) model was employed to simulate wetland landscapes in 2020, the accuracy and limitation of which in simulating lacustrine wetlands were also explored. Lastly, the changing trends of wetland landscapes in 2030 under different development scenarios were predicted. The results show that the landscape of Shengjin Lake Reserve has changed significantly during 2010–2020, with increases in mudflats, reservoirs/ponds, woodlands, and built-up land, and there has been decreases in lakes, grass beaches, and croplands. The PLUS model demonstrated an ideal simulation accuracy for Shengjin Lake Reserve, with the overall accuracy exceeding 80%, kappa coefficient greater than 0.75, and figure of merit (FOM) coefficient of 0.35, indicating that the model can capture the dynamic changes in wetland landscapes accurately. The simulation accuracy can be effectively improved with the adjacent initial year, shorter time interval, and the primary driver factors. Under the natural development scenario, the number of patches in the Shengjin Lake Reserve increased sharply, and landscape fragmentation intensified. Under the urban development scenario, the expansion of built-up land increased, and the average patch area increased. In the ecological protection scenario, the Shannon diversity index and Shannon evenness index of the landscape improved significantly, and the natural wetlands such as grass beaches and lakes can be protected effectively. Our study confirms the applicability of the PLUS model in simulating and predicting lacustrine wetlands landscapes, and the conclusions provide a scientific basis for formulating reasonable development strategies to realize wetland resource conservation and management.

Phycocyanin (PC) concentration is used as an indicator to characterize cyanobacteria biomass while monitoring eutrophication in inland water. Remote sensing provides useful methods for quantifying PC concentration; however, there is a shortage of datasets for the long-term monitoring of PC concentration when only a single remote sensing data is used. Therefore, PC concentrations obtained from multisource remote sensing images should be compared before integrating them for long-term monitoring. In this study, machine learning (ML) regression algorithms are used to develop PC concentration retrieval models suitable for Moderate Resolution Imaging Spectroradiometer (MODIS) and Sentinel-3 Ocean and Land Colour Instrument (OLCI) images, and their accuracies are compared. The two optimal retrieval models are applied to satellite images acquired on the same days to compare the spatial consistency of the two PC concentration retrieval results. The results show that the sensitive spectral range of PC concentration is 560–680 nm. Among the ML regression algorithms, gradient boosted tree (GBT) regression exhibits the highest PC retrieval accuracy for both the MODIS images (R² = 0.82, RMSE = 61.9 μg/L) and OLCI images (R² = 0.86, RMSE = 45.44 μg/L). The PC concentrations retrieved from the MODIS and OLCI images acquired in bloom and no-bloom periods have a high spatial consistency in most areas of Chaohu Lake. Their correlation coefficient also exceeds 0.7, and the average relative error reaches 0.293 μg/L. However, a large difference exists in areas with high PC concentrations, which may cause by the poor applicability of atmospheric correction algorithms and PC retrieval models in these areas. The proposed PC concentration retrieval models developed using GBT regression in this paper can expend the idea for the quantitative retrieval of other inland water quality parameters in inland water, and the conclusions should enable the effective integration of MODIS and OLCI images for the time series monitoring of PC concentrations in reservoirs and lakes.