Oil spill accidents can cause severe ecological disasters; hence, the timely and effective detection of oil spills on the marine surface is of great significance. Synthetic aperture radar (SAR) is very suitable for large-scale oil spill monitoring. As a more advanced form of SAR, polarimetric SAR (PolSAR) can provide more scattering information of land objects, which can help to improve the accuracy of oil spill detection. However, the current studies of oil spill detection by SAR data have mainly focused on using SAR intensity or amplitude information, and the phase information and other polarimetric information have not been fully utilized. To solve this problem, using Sentinel-1 dual-polarimetric images as the data source, this article presents an intelligent oil spill detection architecture based on a deep convolutional neural network (DCNN), in which both the amplitude information and phase information are utilized. Furthermore, to improve the feature discrimination capability, the Cloude polarimetric decomposition parameters are also integrated into the proposed model. The results show that the improved DeepLabv3+ model, which takes ResNet-101 as the backbone network and group normalization (GN) as the normalization layer, can achieve superior performance than those traditional methods. Moreover, the model is better able to capture the fine details of oil spill instances and can achieve fine-scale segmentation.

Dead fuel moisture content (DFMC) is of important significance for estimating and predicting forest wildfire risk, in which 1-h dead fuel is most critical as the easiest fuel to ignite. Current methodologies based on empirical and physical models rely heavily on meteorological data from uneven and sparse stations. In contrast, satellite data become a better choice since its continuous surface observations. Of all satellites, Himawari-8 is the most appropriate data to meet the rapid changes of DFMC throughout the day owing to its high time resolution. Thus, this study explored the application of 1-h DFMC predicting using machine learning from Himawari -8 data. Random forest was selected for the prediction and linear regression was used for comparison. The results showed that random forest has a satisfactory performance with higher R2 (0.53) and lower RMSE (3.15%) than that of linear regression (R2=0.21, RMSE=5.47%). The research suggested that predicting 1-h DFMC at regional scales using machine learning from Himawari -8 data is promising.

Aboveground biomass (AGB) is an important indicator of grassland state. Remote sensing estimation method of grassland biomass can provide important decision support for decision-makers and relevant personnel on grassland management, such as the rational development and utilization of grassland resources, the local ecological environment protection and the rational development of animal husbandry. In this study, based on Landsat 8 and Sentinel-2 A/B satellites data, PROSAILH radiative transfer model (RTM) and look up table (LUT) algorithm were applied to temporally retrieve and map the grassland AGB in Qinghai Province from September 2019 to August 2020, an important graziery region in China. Further analysis on the relation between the AGB and meteorological data showes that the precipitation and air temperature in Qinghai is consistent with the average biomass dynamic, indicating the considerable effect of the meteorological factors on the AGB in this region.

Forest Dead Fuel Load (FDFL) is vital for fire prevention and suppression since it affects the surface fire ignition and intensity directly. The accurate spatial distribution information of FDFL can provide decision support for fire managers. Remote sensing (RS) technology is a unique way to estimate FDFL on a large scale. However, most researches focus on the application of LiDAR data which is expensive to analyze fuel dynamics on spatiotemporal scale. Little attention has been devoted to freely accessible datasets such as Sentinel-1 and Sentinel-2 with global coverage. This study not only combined these two data but also the site conditions (i.e., elevation, slope and aspect) to estimate the FDFL in the southwest of Sichuan, China. The machine learning method, Random Forest Regression (RFR) was selected to manage the multiple and nonlinear relationships between RS data and FDFL. Results show that 1h and all FDFL (the sum of 1h, 10h, 100h and litter) can be more indicated by RS data (1h:$\mathrm{R}^{2}=0.57,\ \text{RMSE}=0.18$Tons/ha; All:$\mathrm{R}^{2}=0.59,\ \text{RMSE}=1.81$Tons/ha). The representation ability of RS data for 10h, 100h and litter is relatively weaker (10h:$\mathrm{R}^{2}=0.41,\ \text{RMSE}=0.47$Tons/ha; 100h:$\mathrm{R}^{2}=0.40,\ \ \text{RMSE}=1.07$Tons/ha; litter:$\mathrm{R}^{2}=0.29,\quad \text{RMSE}=1.66$Tons/ha). Hence, this study demonstrated the potential of multi-source RS data for FDFL estimation.

Wildfire, one of the most serious disasters in the world, making a huge threat to local economic development and residents' life and property safety. Therefore, it is crucial to realize near real time wildfire detection in specific high wildfire risk area. Himawari-8 is a state of art Geostationary Orbit Satellites (GOS), which can realize near-real time wildfire detection and dynamic monitoring of wildfire changes. In this study, three features strategies: spectral, spectral calculated and spectral with spatial were used to extract features from Himawari-8 and Topography data. Next, Random Forest models were trained by these features. And then, we evaluated models on Himawari-8 data generated in March 29, 2020 6:10 UTC, March 30, 2020 5:50 UTC, March 31, 2020 5:30 UTC and April 1,2020 5:50 UTC to detect specific wildfire events in Southwestern China. The results showed an overall precision of 64.24%, 64.08% and 68.07%, and an overall F1-score of 0.737, 0.677, 0.724 based on three different features strategies. This study proved that our models have the ability to detect wildfire point accurately, and model with strategy.3 performed the best when considering precision, omission and F1-score.