Mangroves are coastal vegetation and considered as one of the blue carbon ecosystems. Canopy height is a critical parameter in understanding these ecosystems and related to biomass information. Remote sensing and machine learning techniques are increasingly utilized for large-scale mangrove canopy height mapping. Sentinel-2 is an optical satellite with various bands, including red-edge (RE) bands that are related to vegetation biophysical information. This study investigates the potential of red-edge spectral indices for mangrove canopy height mapping using Random Forest Regression (RFR). The study area is located around Charlotte Harbor Preserve State Park, Florida, USA in 2020. Sentinel-2 data was used to produce several red-edge spectral indices including NDVI, NDVIRE1, NDVIRE2, NDVIRE3, CIRE1, CIRE2, CIRE3, and CIVI which served as inputs for the RFR model. Spaceborne GEDI LiDAR rh98 canopy height data was used here to produce the target data. The dataset was divided into 80% training and 20% testing subsets. Our results indicate that CIVI utilizing red-edge 1 and 2 is the most important feature of RFR for mangrove canopy height estimation. The mean absolute error (MAE) and the root mean squared error (RMSE) based on the testing dataset were 1.662 m and 2.291 m, respectively. To further validate the results, we introduced an independent testing dataset using a canopy height model from the airborne LiDAR data located near the study area. We used the trained RFR model to predict the mangrove canopy height in the testing area and found the MAE and RMSE from the independent testing dataset were 2.511 m and 2.812 m, respectively. Based on the results, most of the red-edge spectral indices have a better performance than the spectral index without red-edge bands, and the use of several red-edge spectral indices provides promising results for mangrove canopy height mapping.

In recent years, advancements in Synthetic Aperture Radar Interferometry (InSAR) technology have accelerated progress in geology and engineering. Particularly in urban areas, using Persistent Scatterer Interferometric Synthetic Aperture Radar (PSInSAR) technology allows for acquiring long-term stable phase information to measure ground deformation down to cm level accuracy. Our research aims to implement PSInSAR technology in Taipei, which has experienced rapid development over the past few decades. However, due to the cutoff of the Keelung River in the 1990s, issues such as ground subsidence and building safety have become focal points of concern. Cracks and foundation subsidence around construction sites further highlight the necessity of monitoring this area. We utilize Sentinel-1 satellite imagery from the European Space Agency to track deformations in 2019-2023 using PSInSAR technology, and we also implement corner reflectors to verify its accuracy. Preliminary results indicate that some patches in our study area exhibit distinct and ongoing subsidence. The maximum subsidence rate is 9 mm per year.

Satellites and high-altitude unmanned aerial vehicles are the top platforms for electro-optical remote sensing for both civilian and military applications. Since early 2000s high altitude electro-optical remote sensing platforms have been actively used for real-time and offline damage assessment following natural disasters such as earthquakes, floods, and landslides. High accuracy, multi-class automated object segmentation is one of the key processing blocks that makes such applications practical. Given the typical distances between target areas and high-altitude sensing platforms (10s to 1000s of kms) as well as the critical nature of the resulting assessments, the accuracy of segmentation maps is of key interest. In this work we present the Multi-Class Certainty Mapped Network (MCCM-Net) that uses multi-class per-pixel uncertainty to enhance segmentation performance. MCCM-Net explicitly models multi-class uncertainty as the entropy of class probability distribution. Pixel-level uncertainty is then used to iteratively enhance segmentation maps. Our experiments on publicly available benchmark datasets show that MCCM-Net provides state-of-the-art multi-class pixel-level segmentation performance.

In the context of increasing climate change and rapid urban expansion in Ho Chi Minh City, rising temperatures pose significant heat stress and health risks to residents. Assessing heat vulnerability is crucial for improving thermal comfort and achieving sustainable urban development. However, there is limited research on heat vulnerability within Local Climate Zones (LCZs), which are essential for understanding the relationship between urban form and heat exposure. This study integrates the Heat Vulnerability Index (HVI) with LCZ analysis to evaluate heat vulnerability across the city. Data were sourced from Landsat 8 to calculate Land Surface Temperature (LST), the Enhanced Normalized Difference Impervious Surfaces Index (ENDISI), the Modified Normalized Difference Water Index (MNDWI), and for LCZ classification using a random forest algorithm; LandScan for population density; OpenStreetMap for road and medical facility density; and VIIRS for nighttime lighting. The HVI model combines exposure (represented by Land Surface Temperature), sensitivity (calculated as an average of population density, road density, and ENDISI), and adaptability (calculated as an average of nighttime lighting, medical facilities, and MNDWI). The results show that built LCZs have significantly higher thermal exposure, sensitivity, and adaptability compared to natural LCZs. Greater vulnerability was observed in urbanized areas, with LCZ1 and LCZ2 having the highest HVI scores, and over 90% of their areas classified as highly vulnerable. Additionally, 158 out of 315 wards had HVI values exceeding 60%, with most located in urban areas. Notably, rural wards such as Pham Van Coi, Cu Chi, and Hoa Phu were identified as particularly vulnerable and in need of immediate attention. In conclusion, integrating LCZ and HVI provides key insights into heat vulnerability patterns, helping future studies optimize the spatial layouts of local climate zones to mitigate heat impacts and improve the urban thermal environment For the keywords, select up to 8 key terms for a search on your manuscript's subject.

Semantic segmentation of urban areas is crucial for many applications. Self-supervised networks require few or no labels for training, making them highly appealing approaches. One such network is STEGO¹, which builds upon DINO² and operates without any labeled data, yet effectively segments buildings, vegetation, and roads in the ISPRS Potsdam dataset³. The resulting segmentations are refined using Conditional Random Fields (CRF). In remote sensing, additional channels like the Normalized Digital Surface Model (NDSM) enhance the segmentation task, as pixels of the same class often exhibit similar elevation characteristics, especially when adjacent. Since the transformer-based DINO network is built for RGB data, we extend the CRF with NDSM information to overcome this limitation, introducing a second pairwise potential that encourages neighboring pixels with similar elevation to have the same label. For evaluation in both the linear and cluster probe, we employ Adjusted Mutual Information (AMI) and Adjusted Rand Index (ARI) to assess the segmentation against the six classes of the Potsdam dataset, besides standard IoU and accuracy metrics. Enhancing the CRF with elevation information improves the mIoU by 0.83% over the RGB only baseline in the cluster probe, which constitutes a considerable improvement.

Assuming a linear relationship between shortwave infrared (SWIR) and near-infrared (VNIR), it may be possible to estimate ground information obscured by smoke in the VNIR using SWIR. This study applies the least squares method for linear regression analysis, using the Greek wildfire on August 23, 2023, and simulated imagery as case studies to demonstrate the feasibility of this approach.

Climate change causes weather pattern changes and heavy rainfall more frequently which causes flooding in many areas. Due to its limitations on cloud cover, optical satellites are not capable of mapping floods. SAR satellites such as Sentinel-I SAR is a high-resolution satellite used for detecting flood inundation because of its capability to penetrate clouds and not depend on the weather. The Otsu thresholding algorithm was applied to identify an optimal threshold for each preprocessed Sentinel-I image to separate water from non-water pixels for producing the best threshold to identify flood and non-flood areas based on the backscatter intensity. The results indicate Sentinel-I SAR shows that the application of VH polarization with the Otsu method can map floods well in Central Java flood events. For the impact of flooding, ESA WorldCover with 10m resolution is used to estimate the amount of affected cropland and urban areas. The population density is obtained from The Global Human Settlement Layer (GHSL) to count the people exposed to flood. In the March 15 at the early stage of the flood event, 63,698 hectares were estimated affected, and the number of exposed people was 28,000. Based on ESA WorldCover data, 44,080 hectares of cropland are affected and 1,119 hectares of urban areas are affected by flood events. With the rapid mapping process of disaster-affected areas, it will be easier for the decision-makers and disaster management to make proper decisions during the disaster time.

The 2024 Noto Peninsula earthquake occurred on January 1 with a series of Mj5.5 and Mj7.6 earthquakes. The tsunami triggered by this earthquake simultaneously caused extensive damage to the coastal areas of the Noto Peninsula. In addition, buildings collapsed over a wide area in Wajima City. Furthermore, fires caused by these earthquakes resulted in the largescale destruction of buildings. In the case of such a large-scale damage, it is necessary to grasp the damage situation as soon as possible, and it is effective to collect damage information by satellite remote sensing. Furthermore, the resolution of the satellite image has been increased, and Method of object-based and pixel-based texture for extracting detailed damage information of the building and grasping the damaged area have been studied. The damage caused by the optical and SAR images observed immediately after the damage has been investigated and the damage of the building by the field survey is being investigated. However, there are few examples of studies that have applied texture analysis of satellite imagery to include building destruction by fire and building collapse. In this study, the distribution characteristics of texture indices using high-resolution optical satellite images observed before and after the disaster in Wajima City were evaluated based on the results of the field survey. Those results show the characteristics of texture indicators for building damage, including burnout. Furthermore, the trend of change in the texture index effectively discussed the applicability of the disaster in building damage.

The Choushui River alluvial fan is an important agricultural production region in western Taiwan. This insufficient level of rainfall replenishment, combined with the unauthorized use of wells to irrigate crops, exacerbated land subsidence in the region. In this study, leveling, global navigation satellite system (GNSS), and interferometric synthetic aperture radar (InSAR) observation data were collected from 2018 to 2021 for analysis. The GNSS data were processed using the precise point positioning method, and the results were compared with data processed using the traditional static relative positioning method. The three-year cumulative subsidence at the subsidence center was approximately 9 cm in Changhua and 19 cm in Yunlin. To compensate for the lack of spatial resolution of GNSS and improve the accuracy of InSAR, the InSAR, GNSS, and leveling data were integrated, and these results were validated against leveling check points. They showed an average error of 0.4 cm/year in the annual subsidence rate. Overall, the GNSS had high precision and continuity, making it suitable for real-time early warning of land subsidence.

Efficient analysis of ground cover types is vital during the investigative stages of civil construction projects. Traditional surveying methods, although accurate, are often expensive, time-consuming, and cumbersome, especially for large-scale infrastructure developments. This study explores the way of enhancing the automation of ground cover classification using high-resolution UAV imagery and machine learning. Various classification algorithms, including Support Vector Machine (SVM) and k-Nearest Neighbour (kNN), were tested. Post-processing technique such as multi-image averaging was employed to mitigate the effects of shadows and transient objects like vehicles, resulting in improved classification accuracy. Accuracy assessments comparing UAV-based automated classification results with manually classified datasets and feature survey data demonstrate the potential of this approach to streamline data capture during the early stages of civil construction projects. The use of UAVs not only reduces costs but also accelerates the data collection process. However, challenges such as noise, shadows, and misclassifications persist, indicating the need for further refinement and integration of additional data sources like multispectral imagery and UAV LiDAR data. The findings suggest that automated ground cover classification can support quicker decision-making in civil construction, improving the efficiency of future data collection and planning processes. The SVM algorithm proved to be the more effective method, consistently achieving higher accuracy in classifying the raster datasets into the designated classes. This study contributes to improved decision-making in construction planning and highlights the potential for further advancements in automated classification technologies in construction environments.

In recent years, satellite imagery has become a very useful tool for managing disasters and protecting the environment. When disaster happens, satellites can give reliable information about how serious is the disaster and where the area that are affected. Satellite imagery covers extensive surface areas and provides near real-time data, it helps agencies access remote or inaccessible regions. In this study, we focus on earthquakes that hit Ishikawa in Japan and eastern island of Taiwan earlier this year. When disaster occurred, FORMOSAT-5 responds quickly for providing imaging and value add products. Additionally, we also obtained valuable satellite data through collaborated with sentinel Asia, including Japan, Thailand and India. To identify the impacted area, we compared pre and post scenes with visualize change analysis rapidly to provide availability of information for decision-making. Soon afterwards, Numerical analysis was also provided. As the result, we were able to identify buildings were damaged in Wajima City and new exposed land extending along with coastline, on the other hand, large-scale landslides and a barrier lake have been discovered in mountainous areas of Taiwan.