Explore the vast range of titles available.

Floods are the second‐most deadly natural hazard in Australia, following heatwaves. Monitoring flood extent and depth in near real‐time (NRT) is crucial to minimize loss of life and socio‐economic impacts. This study leverages advanced computing, data management systems, and high‐quality data, including river gauge data APIs and Australian Water Outlook, Digital Earth Australia, Google Earth Engine and Amazon Web Service, to develop a flood monitoring workflow in Australia. Our framework provides NRT 5‐m spatial resolution flood extent and depth maps using airborne LiDAR observations through three approaches: (a) gauge data, (b) coupled hydrological and hydrodynamics model, and (c) satellite observations (i.e., Sentinel‐1, Sentinel‐2, Landsat‐7/8/9). We evaluated this flood monitoring framework in seven river catchments across Australia, using both deterministic and ensemble modes. This study highlights the importance of low‐latency gauge data for flood monitoring, as well as the necessity of high‐resolution airborne LiDAR DEMs for accurate flood mapping. In ungauged areas, the ensemble modeling approach enhances the model's ability to capture flood inundation dynamics. In cases where this remains challenging, multi‐source remote sensing can help mitigate the limitations of the modeling approach. We also demonstrated the potential for transferring this flood monitoring framework to other regions around the world. Overall, this study advances the operationalization of high‐resolution flood analytics, offering a replicable blueprint to strengthen community resilience against escalating flood risks under climate change.

We performed a comparative analysis of the prediction accuracy of machine learning methods and ordinary Kriging (OK) hybrid methods for forest volume models based on multi-source remote sensing data combined with ground survey data. Taking Larix olgensis, Pinus koraiensis, and Pinus sylvestris plantations in Mengjiagang forest farms as the research object, based on the Chinese Academy of Forestry LiDAR, charge-coupled device, and hyperspectral (CAF-LiTCHy) integrated system, we extracted the visible vegetation index, texture features, terrain factors, and point cloud feature variables, respectively. Random forest (RF), support vector regression (SVR), and an artificial neural network (ANN) were used to estimate forest volume. In the small-scale space, the estimation of sample plot volume is influenced by the surrounding environment as well as the neighboring observed data. Based on the residuals of these three machine learning models, OK interpolation was applied to construct new hybrid forest volume estimation models called random forest Kriging (RFK), support vector machines for regression Kriging (SVRK), and artificial neural network Kriging (ANNK). The six estimation models of forest volume were tested using the leave-one-out (Loo) cross-validation method. The prediction accuracies of these six models are better, with RLoo2 values above 0.6, and the prediction accuracy values of the hybrid models are all improved to different extents. Among the six models, the RFK hybrid model had the best prediction effect, with an RLoo2 reaching 0.915. Therefore, the machine learning method based on multi-source remote sensing factors is useful for forest volume estimation; in particular, the hybrid model constructed by combining machine learning and the OK method greatly improved the accuracy of forest volume estimation, which, thus, provides a fast and effective method for the remote sensing inversion estimation of forest volume and facilitates the management of forest resources.

Alpine peatlands are one of the carbon reservoirs, provide vital ecosystem services, and support endangered biodiversity. However, they are globally understudied, including those in the Italian Alps, which host thousands of small sites averaging under 1 ha. Their complex geomorphology makes detection challenging with single‐sensor, low‐resolution remote sensing imagery. In the last decade, high resolution multi‐source imagery (e.g., Sentinel series) and the cloud‐based computation platforms (e.g., Google Earth Engine—GEE) have become available. Using these advancements, we developed a method to map the distribution of alpine peatlands. Utilizing 1 and 30 m digital elevation models (DEMs), optical, and microwave data sets, our method exploits a pixel‐based Random Forest (RF) machine‐learning algorithm on the GEE platform to map alpine peatlands in the Avisio River basin of the Italian Alps. The results show that the data set of single‐year time series multi‐source imagery, binary samples (peatland or non‐peatland), and 30 m DEM is the most effective for mapping alpine peatlands. The method achieved an overall accuracy of 90.5%, with 81.8% true positives and 0.8% false positives. The method identified 11.635 km2 of alpine peatlands, surpassing the 7.764 km2 documented in official inventories, this discrepancy may be due to overestimation but also gaps in the existing reference inventory. In the classification process, DEM derived variables proved more effective than optical and microwave derived variables. Variable importance analysis in the RF model indicated that elevation is the most influential factor, while the microwave derived VV‐VH difference (ascending track) contributes the least. Plain Language Summary Alpine peatlands are globally underestimated, highlighting the need for effective and regularly updated mapping methods to better understand their distribution and extent. With advancements in remote sensing technology, large‐scale mapping of alpine peatlands using satellite imagery has become increasingly feasible. While previous studies have achieved notable results using 30 m resolution satellite imagery, the performance and potential of 10 m resolution imagery have not been thoroughly validated. This study focused on mapping alpine peatlands within a specific river basin in the Italian Alps. We developed a mapping approach using 10 m resolution multi‐source time‐series remote sensing data from Sentinel‐1 (microwave) and Sentinel‐2 (optical) sensors, processed on the cloud‐based Google Earth Engine platform, and applied a Random Forest image classification algorithm. The method demonstrated a high overall accuracy of 90.5%. Our findings indicate that variables derived from digital elevation models were more influential in the classification process than those derived from optical and microwave imagery. The alpine peatlands classification method can be extended to other regions, providing a valuable reference for mapping peatlands both within the Alps and in other similar environments. Key Points Using Random Forest to map alpine peatlands is effective when optical and microwave data are combined with hydrogeomorphological indices Alpine topography and geomorphology contribute more significantly to peatland identification than optical and microwave data The method achieves high overall accuracy, even the small size of peatlands and the mountainous topography facilitate the overestimation

Accurately estimating large-area crop yields, especially for soybeans, is essential for addressing global food security challenges. This study introduces a deep learning framework that focuses on precise county-level soybean yield estimation in the United States. It utilizes a wide range of multi-variable remote sensing data. The model used in this study is a state-of-the-art CNN-BiGRU model, which is enhanced by the GOA and a novel attention mechanism (GCBA). This model excels in handling intricate time series and diverse remote sensing datasets. Compared to five leading machine learning and deep learning models, our GCBA model demonstrates superior performance, particularly in the 2019 and 2020 evaluations, achieving remarkable R 2 , RMSE, MAE and MAPE values. This sets a new benchmark in yield estimation accuracy. Importantly, the study highlights the significance of integrating multi-source remote sensing data. It reveals that synthesizing information from various sensors and incorporating photosynthesis-related parameters significantly enhances yield estimation precision. These advancements not only provide transformative insights for precision agricultural management but also establish a solid scientific foundation for informed decision-making in global agricultural production and food security.

[This corrects the article DOI: 10.3389/fpls.2025.1730583.].

With the increasingly severe damage wreaked by forest fires, their scientific and effective prevention and control has attracted the attention of countries worldwide. The breakthrough of remote sensing technologies implemented in the monitoring of fire spread and early warning has become the development direction for their prevention and control. However, a single remote sensing data collection point cannot simultaneously meet the temporal and spatial resolution requirements of fire spread monitoring. This can significantly affect the efficiency and timeliness of fire spread monitoring. This article focuses on the mountain fires that occurred in Muli County, on 28 March 2020, and in Jingjiu Township on 30 March 2020, in Liangshan Prefecture, Sichuan Province, as its research objects. Multi-source satellite remote sensing image data from Planet, Sentinel-2, MODIS, GF-1, GF-4, and Landsat-8 were used for fire monitoring. The spread of the fire time series was effectively and quickly obtained using the remote sensing data at various times. Fireline information and fire severity were extracted based on the calculated differenced normalized burn ratio (dNBR). This study collected the meteorological, terrain, combustibles, and human factors related to the fire. The random forest algorithm analyzed the collected data and identified the main factors, with their order of importance, that affected the spread of the two selected forest fires in Sichuan Province. Finally, the vegetation coverage before and after the fire was calculated, and the relationship between the vegetation coverage and the fire severity was analyzed. The results showed that the multi-source satellite remote sensing images can be utilized and implemented for time-evolving forest fires, enabling forest managers and firefighting agencies to plan improved firefighting actions in a timely manner and increase the effectiveness of firefighting strategies. For the forest fires in Sichuan Province studied here, the meteorological factors had the most significant impact on their spread compared with other forest fire factors. Among all variables, relative humidity was the most crucial factor affecting the spread of forest fires. The linear regression results showed that the vegetation coverage and dNBR were significantly correlated before and after the fire. The vegetation coverage recovery effects were different in the fire burned areas depending on fire severity. High vegetation recovery was associated with low-intensity burned areas. By combining the remote sensing data obtained by multi-source remote sensing satellites, accurate and macro dynamic monitoring and quantitative analysis of wildfires can be carried out. The study’s results provide effective information on the fires in Sichuan Province and can be used as a technical reference for fire spread monitoring and analysis through remote sensing, enabling accelerated emergency responses.

Harmful algal blooms (hereafter HABs) pose significant threats to aquatic health and environmental safety. Although satellite remote sensing can monitor HABs at a large-scale, it is always a challenge to achieve both high spatial and high temporal resolution simultaneously with a single earth observation system (EOS) sensor, which is much needed for aquatic environment monitoring of inland lakes. This study proposes a multi-source remote sensing-based approach for HAB monitoring in Chaohu Lake, China, which integrates Terra/Aqua MODIS, Landsat 8 OLI, and Sentinel-2A/B MSI to attain high temporal and spatial resolution observations. According to the absorption characteristics and fluorescence peaks of HABs on remote sensing reflectance, the normalized difference vegetation index (NDVI) algorithm for MODIS, the floating algae index (FAI) and NDVI combined algorithm for Landsat 8, and the NDVI and chlorophyll reflection peak intensity index (ρchl) algorithm for Sentinel-2A/B MSI are used to extract HAB. The accuracies of the normalized difference vegetation index (NDVI), floating algae index (FAI), and chlorophyll reflection peak intensity index (ρchl) are 96.1%, 95.6%, and 93.8% with the RMSE values of 4.52, 2.43, 2.58 km2, respectively. The combination of NDVI and ρchl can effectively avoid misidentification of water and algae mixed pixels. Results revealed that the HAB in Chaohu Lake breaks out from May to November; peaks in June, July, and August; and more frequently occurs in the western region. Analysis of the HAB’s potential driving forces, including environmental and meteorological factors of temperature, rainfall, sunshine hours, and wind, indicated that higher temperatures and light rain favored this HAB. Wind is the primary factor in boosting the HAB’s growth, and the variation of a HAB’s surface in two days can reach up to 24.61%. Multi-source remote sensing provides higher observation frequency and more detailed spatial information on a HAB, particularly the HAB’s long-short term changes in their area.

Leaf Area Index (LAI) is an important parameter which can be used for crop growth monitoring and yield estimation. Many studies have been carried out to estimate LAI with remote sensing data obtained by sensors mounted on Unmanned Aerial Vehicles (UAVs) in major crops; however, most of the studies used only a single type of sensor, and the comparative study of different sensors and sensor combinations in the model construction of LAI was rarely reported, especially in soybean. In this study, three types of sensors, i.e., hyperspectral, multispectral, and LiDAR, were used to collect remote sensing data at three growth stages in soybean. Six typical machine learning algorithms, including Unary Linear Regression (ULR), Multiple Linear Regression (MLR), Random Forest (RF), eXtreme Gradient Boosting (XGBoost), Support Vector Machine (SVM) and Back Propagation (BP), were used to construct prediction models of LAI. The results indicated that the hyperspectral and LiDAR data did not significantly improve the prediction accuracy of LAI. Comparison of different sensors and sensor combinations showed that the fusion of the hyperspectral and multispectral data could significantly improve the predictive ability of the models, and among all the prediction models constructed by different algorithms, the prediction model built by XGBoost based on multimodal data showed the best performance. Comparison of the models for different growth stages showed that the XGBoost-LAI model for the flowering stage and the universal models of the XGBoost-LAI and RF-LAI for three growth stages showed the best performances. The results of this study might provide some ideas for the accurate estimation of LAI, and also provide novel insights toward high-throughput phenotyping of soybean with multi-modal remote sensing data.

Vegetation serves as the most critical carbon reservoir within terrestrial ecosystems and plays a vital role in mitigating global climate change. Australia features a vast and diverse landscape, ranging from dense eucalyptus forests to sparse woodlands, and harbors rich biodiversity. However, the significant spatial heterogeneity across the continent presents substantial challenges for accurately estimating regional aboveground biomass (AGB). This study aims to assess the accuracy of various models in AGB estimation. The dataset includes field-measured biomass and multi-source remote sensing data, such as vegetation canopy height products, Landsat imagery, topographic data, and climate variables. To build biomass estimation models, a Stacking regressor is constructed, and extensive comparative experiments were conducted. The Stacking model comprises seven base learners and one meta-learner. The meta-learner learns to optimally combine the predictions of the base models by minimizing prediction error. The experiments include: (1) comparing the performance of the Stacking model with seven individual machine learning models using K-fold cross-validation, and (2) evaluating the impact of recursive feature elimination (RFE) on model performance before and after feature selection. Based on predictive performance, the Stacking, Gradient Boosting Regressor (GBR), and Random Forest (RF) models are selected for biomass mapping. In addition, we introduced the Monte Carlo simulation approach to evaluate the uncertainty of biomass estimation. Results show that the Stacking model outperforms all individual machine learning models across all forest types. The final values for the Stacking, GBR, and RF models are 0.74, 0.72, and 0.71, respectively, with RMSE values of 49.79 Mg/ha, 51.00 Mg/ha, and 52.18 Mg/ha. The integration of multi-source data effectively addresses the limitations of single-source datasets, and the ensemble learning approach enhances the robustness of biomass estimation. This study provides theoretical and technical support for large-scale AGB estimation in heterogeneous vegetation landscapes.

On August 12, 2015, a catastrophic landslide occurred in Shanyang County, Shaanxi Province, China, resulting in 7 deaths and 53 missing. This study investigates the lifecycle evolution and failure mechanism of the Shanyang landslide with multi-source remote sensing data, emphasizing the critical role of locked segments in the Shanyang landslide. Differential interferometric analysis and deformation decomposition were utilized to reveal the pre-failure deformation pattern of the Shanyang landslide. Creeping deformation was found along the underlying soft layer 4 months prior to the landslide, with the deformation mainly occurring downslope and controlled by the locked segment at the front edge of the slope. The integration of a 1:1000 pre-failure topographic map and a high-precision post-failure digital elevation model determined the landslide volume to be 1.60 × 106 m3 and revealed a maximum travel distance of 500 m. Combining engineering geological zoning with deformation data, the Shanyang landslide was classified as a typical locked-segment-dominated slide in soft-hard interbedded strata, with rainfall as a key deformation influence factor. Finally, using the time series deformation from SBAS-InSAR, the post-failure stability of the landslide area was analyzed. This study demonstrates the potential of integrating multi-temporal remote sensing techniques to identify the entire deformation and destruction process of landslides and their influencing factors, which offers valuable insights for improving early landslide warnings and hazard assessments.