Owning to the nature of flood events, near-real-time flood detection and mapping is essential for disaster prevention, relief, and mitigation. In recent years, the rapid advancement of deep learning has brought endless possibilities to the field of flood detection. However, deep learning relies heavily on training samples and the availability of high-quality flood datasets is rather limited. The present study collected 16 flood events in the Yangtze River Basin and divided them into three categories for different purpose: training, testing, and application. An efficient methodology of dataset-generation for training, testing, and application was proposed. Eight flood events were used to generate strong label datasets with 5296 tiles as flood training samples along with two testing datasets. The performances of several classic convolutional neural network models were evaluated with those obtained datasets, and the results suggested that the efficiencies and accuracies of convolutional neural network models were obviously higher than that of the threshold method. The effects of VH polarization, VV polarization, and the involvement of auxiliary DEM on flood detection were investigated, which indicated that VH polarization was more conducive to flood detection, while the involvement of DEM has a limited effect on flood detection in the Yangtze River Basin. Convolutional neural network trained by strong datasets were used in near-real-time flood detection and mapping for the remaining eight flood events, and weak label datasets were generated to expand the flood training samples to evaluate the possible effects on deep learning models in terms of flood detection and mapping. The experiments obtained conclusions consistent with those previously made on experiments with strong datasets.

Gross Ecosystem Product (GEP) is a single currency to measure the contribution of an ecosystem to society. It is the sum of the value of products and services provided by the ecosystem to humans. Based on this, GEP characterizes the current situation of the ecological environment in a region and can measure the contribution index provided by ecosystems. Based on land use data, rainfall data, soil data, statistical yearbook, DEM and other data, this paper constructs the accounting framework of Hainan Island’s GEP, quantitatively evaluates Hainan Island’s GEP from 2010 to 2020 and analyzes the research results. The results are as follows: (1) The GEP of Hainan Island’s ecosystem rose from 596.404 billion CNY in 2010 to 1032.096 billion CNY in 2020, an increase of about 42.21% with an average annual growth rate of 6.3%. The overall GEP has shown an upward trend. Among them, the value of regulation service accounts for the largest proportion was up to 73–83%. (2) The overall distribution of GEP in Hainan Island is uneven, decreasing from the central mountainous areas to the eastern coastal areas around to the west. In particular, Danzhou City, Wenchang City and Haikou City have higher values, while Tunchang, Baoting County and Wuzhishan City have lower values, and Wuzhishan City has the lowest values. Hainan Island is rich in natural resources, which contain great value. Through the accounting of GEP, we can clearly determine the value of the ecosystem, which can help the relevant departments estimate the GEP of Hainan Island’s ecosystem. It can provide a reference for Hainan to formulate ecological protection and sustainable development policies, and for the GEP estimation of other islands.

Time series Autoregressive Integrated Moving Average (ARIMA) model is often used in landslide prediction and forecasting. However, few conditions have been suggested for the application of ARIMA models in landslide displacement prediction. This paper summarizes the distribution law of the tangential angle in different time periods and analyzes the landslide displacement data by combining wavelet transform. It proposes an applicable condition for the ARIMA model in the field of landslide prediction: when the landslide deformation is in the initial deformation to initial acceleration stage, i.e., the tangential angle of landslide displacement is less than 80°, the ARIMA model has higher prediction accuracy for 24-h landslide displacement data. The prediction results are RMSE = 4.52 mm and MAPE = 2.39%, and the prediction error increases gradually with time. Meanwhile, the ARIMA model was used to predict the 24-h displacements from initial deformation to initial acceleration deformation for the landslide in Guangna Township and the landslide in Libian Gully, and the prediction results were RMSE = 1.24 mm, MAPE = 1.34% and RMSE = 5.43 mm, MAPE = 1.67%, which still maintained high accuracy and thus verified this applicable condition. At the same time, taking the landslide of Libian Gully as an example, the ARIMA model was used to test the displacement prediction effect of the landslide in the Medium-term acceleration stage and the Imminent sliding stage (the tangential angle of landslide displacement is 80° and 85°, respectively). The relative error of displacement data prediction in the Medium-term acceleration stage is within 3%, while the relative error of the prediction value in the Imminent sliding stage is more than 3%, and the error gradually increases with time. This demonstrates that the relative error of the ARIMA model in landslide prediction and forecasting is within 3%. The relative error of the prediction value in the Imminent sliding stage is above 3%, and the error increases gradually with time. Meanwhile, the prediction results are analyzed and it is concluded that the increase in prediction time and tangential angles are the main reasons for the increase in error. The applicable conditions proposed in this study can provide a reference for the application of ARIMA model in landslide prediction and forecast.

Current road extraction models from remote sensing images based on deep learning are computationally demanding and memory-intensive because of their high model complexity, making them impractical for mobile devices. This study aimed to develop a lightweight and accurate road extraction model, called Road-MobileSeg, to address the problem of automatically extracting roads from remote sensing images on mobile devices. The Road-MobileFormer was designed as the backbone structure of Road-MobileSeg. In the Road-MobileFormer, the Coordinate Attention Module was incorporated to encode both channel relationships and long-range dependencies with precise position information for the purpose of enhancing the accuracy of road extraction. Additionally, the Micro Token Pyramid Module was introduced to decrease the number of parameters and computations required by the model, rendering it more lightweight. Moreover, three model structures, namely Road-MobileSeg-Tiny, Road-MobileSeg-Small, and Road-MobileSeg-Base, which share a common foundational structure but differ in the quantity of parameters and computations, were developed. These models varied in complexity and were available for use on mobile devices with different memory capacities and computing power. The experimental results demonstrate that the proposed models outperform the compared typical models in terms of accuracy, lightweight structure, and latency and achieve high accuracy and low latency on mobile devices. This indicates that the models that integrate with the Coordinate Attention Module and the Micro Token Pyramid Module surpass the limitations of current research and are suitable for road extraction from remote sensing images on mobile devices.

As an environmental parameter, the chlorophyll-a concentration (Chl-a) is essential for monitoring water quality and managing the marine ecosystem. However, current mainstream Chl-a inversion algorithms have limited accuracy and poor spatial and temporal generalization in Case II waters. In this study, we constructed a quantitative model for retrieving the spatial and temporal distribution of Chl-a in the Bohai–Yellow Sea area using Geostationary Ocean Color Imager (GOCI) spectral remote sensing reflectance ($R_{rs}\left(\lambda \right)$) products. Firstly, the GOCI $R_{rs}\left(\lambda \right)$ correction model based on measured spectral data was proposed and evaluated. Then, the feature variables of the band combinations with the highest correlation with Chl-a were selected. Subsequently, Chl-a inversion models were developed using three empirical ocean color algorithms (OC4, OC5, and YOC) and four machine learning methods: BP neural network (BPNN), random forest (RF), AdaBoost, and support vector regression (SVR). The retrieval results showed that the machine learning methods were much more accurate than the empirical algorithms and that the RF model retrieved Chl-a with the best performance and the highest prediction accuracy, with a determination coefficient R² of 0.916, a root mean square error (RMSE) of 0.212 mg$·$m⁻³, and a mean absolute percentage error (MAPE) of 14.27%. Finally, the Chl-a distribution in the Bohai–Yellow Sea using the selected RF model was derived and analyzed. Spatially, Chl-a was high in the Bohai Sea, including in Laizhou Bay, Bohai Bay, and Liaodong Bay, with a value higher than 4 mg$·$m⁻³. Chl-a in the Bohai Strait and northern Yellow Sea was relatively low, with a value of less than 3 mg$·$m⁻³. Temporally, the inversion results showed that Chl-a was considerably higher in winter and spring compared to autumn and summer. Diurnal variation retrieval effectively demonstrated GOCI’s potential as a capable tool for monitoring intraday changes in chlorophyll-a concentrations.

The Taklamakan Desert Region (TDR) and the Gobi Desert Region (GDR) in East Asia significantly impact air quality, human health, and climate through dust aerosols. Utilizing the MERRA-2 dataset’s long-term dust aerosol optical depth (DAOD) at 550 nm from 2000 to 2022, we systematically monitored the spatiotemporal dynamics of DAOD. Our analysis covered annual, seasonal, and monthly scales, employing geographical detector analyses to investigate the impact of eight factors on DAOD distribution. Over the 23-year period, the interannual variability in DAOD across East Asia was not pronounced, but a discernible decreasing trend was observed, averaging an annual decrease of −0.0002. The TDR had higher DAOD values (0.337) than the GDR (0.103). The TDR showed an average annual increase of 0.004, while the GDR exhibited an average annual decrease of −0.0003. The spatial distribution displayed significant seasonal variations, with peak values in spring, although the peak months varied between the TDR and GDR. The driving factor analysis revealed that relative humidity and soil moisture significantly impacted the DAOD spatial distribution in East Asia, which were identified as common driving factors for both the region and the major dust sources. Complex mechanisms influenced the variation in DAOD, with interactions between variables having a greater impact than individual effects. The geodetector-derived interaction q-value identified the collective impact of soil temperature and relative humidity (0.896) as having the highest impact on the spatial and temporal DAOD distribution. The overall spatial pattern exhibited a nonlinear enhancement trend, with the TDR and GDR showing bilinear enhancement patterns. These findings contribute to a better understanding of the factors influencing DAOD, offering a theoretical basis for atmospheric pollution control in East Asia.