The outbreak of Pine Shoot Beetle (PSB, Tomicus spp.) posed a significant threat to the health of Yunnan pine forests, necessitating the development of an efficient and accurate remote sensing monitoring method. The integration of unmanned aerial vehicle (UAV) imagery and deep learning algorithms shows great potential for monitoring forest-damaged trees. Previous studies have utilized various deep learning semantic segmentation models for identifying damaged trees in forested areas; however, these approaches were constrained by limited accuracy and misclassification issues, particularly in complex forest backgrounds. This study evaluated the performance of five semantic segmentation models in identifying PSB-damaged trees (UNet, UNet++, PAN, DeepLabV3+ and FPN). Experimental results showed that the FPN model outperformed the others in terms of segmentation precision (0.8341), F1 score (0.8352), IoU (0.7239), mIoU (0.7185) and validation accuracy (0.9687). Under the pure Yunnan pine background, the FPN model demonstrated the best segmentation performance, followed by mixed grassland-Yunnan pine backgrounds. Its performance was the poorest in mixed bare soil-Yunnan pine background. Notably, even under this challenging background, FPN still effectively identified diseased trees, with only a 1.7% reduction in precision compared to the pure Yunnan pine background (0.9892). The proposed method in this study contributed to the rapid and accurate monitoring of PSB-damaged trees, providing valuable technical support for the prevention and management of PSB.

Flood hazards pose a significant threat to communities and ecosystems alike. Triggered by various factors such as heavy rainfall, storm surges, or rapid snowmelt, floods can wreak havoc by inundating low-lying areas and overwhelming infrastructure systems. Understanding the feedback between local geomorphology and sediment transport dynamics in terms of the extent and evolution of flood-related damage is necessary to build a system-level description of flood hazard. In this research, we present a multispectral imagery-based approach to broadly map sediment classes and how their spatial extent and relocation can be monitored. The methodology is developed and tested using data collected in the Ahr Valley in Germany during post-disaster reconnaissance of the July 2021 Western European flooding. Using uncrewed aerial vehicle-borne multispectral imagery calibrated with laboratory-based soil characterization, we illustrate how fine and coarse-grained sediments can be broadly identified and mapped to interpret their transport behavior during flood events and their role regarding flood impacts on infrastructure systems. The methodology is also applied to data from the 2022 flooding of the Yellowstone River, Gardiner, Montana, in the United States to illustrate the transferability of the developed approach across environments. Here, we show how the distribution of soil classes can be mapped remotely and rapidly, and how this facilitates understanding their influence on local flow patterns to induce bridge abutment scour. The limitations and potential expansions to the approach are also discussed.

Landslides can cause severe damage to property and human life. Identifying their locations and characteristics is crucial for emergency rescue and disaster risk assessment. However, existing methods need help in accurately detecting landslides because of their diverse characteristics and scales, as well as the differences in spectral features and spatial heterogeneity of remote sensing images. To overcome these challenges, a multiscale feature fusion landslide-detection network (MFLD-Net) is proposed. This network utilizes reflectance difference images from pre- and post-landslide Sentinel-2A images, along with digital elevation model (DEM) data. Moreover, a multichannel differential landslide dataset was constructed through spectral analysis of Sentinel-2A images, which facilitates network training and enables differentiation between landslides and other objects with similar spectral features, such as bare soil and buildings. The proposed MFLD-Net was tested in Shuzheng Valley and Detuo town in Sichuan, China, where earthquakes have occurred. The experimental results revealed that compared with advanced deep learning models, MFLD-Net has promising landslide detection performance. This study provides suggestions for selecting optimal deep learning methods and spectral band combinations for landslide detection and offers a publicly available landslide dataset for further research.

The mechanical properties of faecal sludge (FS) influence its moisture retention characteristics to a greater extent than other properties. A comprehensive fundamental characterisation of the mechanical properties is scarcely discussed in the literature. This research focused on bulk and true densities, porosity, particle size distribution and zeta-potential, extracellular polymeric substances, rheology and dilatancy, microstructure analysis, and compactibility in the context of using the FS as a substitute for soil in land reclamation and bioremediation processes. FSs from different on-site sanitation systems were collected from around Durban, South Africa. The porosity of the FSs varied between 42% and 63%, with the zeta-potential being negative, below 10 mV. Over 95% of the particles were <1000 m. With its presence in the inner part of the solid particles, tightly bound extra-cellular polymeric substances (TB-EPSs) influenced the stability of the sludge by tightly attaching to the cell walls, with the highest being in the septic tank with the greywater sample. More proteins than carbohydrates also confirmed characterised the anaerobic nature of the sludge. The results of the textural properties using a penetrometer showed that the initial slope of the positive part of the penetration curve was related to the stiffness of the sludge sample and similar to that of sewage sludge. The dynamic oscillatory measurements exhibited a firm gel-like behaviour with a linear viscoelastic behaviour of the sludges due to the change in EPSs because of anaerobicity. The high-TS samples exhibited the role of moisture as a lubricating agent on the motion of solid particles, leading to dilatancy with reduced moisture, where the yield stress was no longer associated with the viscous forces but with the frictional contacts of solid-solid particle interactions. The filtration-compression cell test showed good compactibility, but the presence of unbound moisture even at a high pressure of 300 kPa meant that not all unbound moisture was easily removable. The moisture retention behaviour of FS was influenced by its mechanical properties, and any interventional changes to these properties can result in the release of the bound moisture of FS.

Glacial responses to climate change exhibit considerable heterogeneity. Although global glaciers are generally thinning and retreat, glaciers in the Karakoram region are distinct in their surging or advancing, exhibiting nearly zero or positive mass balance-a phenomenon known as the Karakoram Anomaly. This anomaly has sparked significant scientific interest, prompting extensive research into glacier anomalies. However, the dynamics of the Karakoram anomaly, particularly its evolution and persistence, remain insufficiently explored. In this study, we employed Landsat reflectance data and Moderate Resolution Imaging Spectroradiometer (MODIS) MCD43A3 albedo products to developed high-resolution albedo retrieval models using two machine learning (ML) regressions--random forest regression (RFR) and back-propagation neural network regression (BPNNR). The optimal BPNNR model (Pearson correlation coefficient [r] = 0.77-0.97, unbiased root mean squared error [ubRMSE] = 0.056-0.077, RMSE = 0.055-0.168, Bias = -0.149 similar to -0.001) was implemented on the Google Earth Engine cloud-based platform to estimate summer albedo at a 30-m resolution for the Karakoram region from 1990 to 2021. Validation against in-situ albedo measurements on three glaciers (Batura, Mulungutti and Yala Glacier) demonstrated that the model achieved an average ubRMSE of 0.069 (p < 0.001), with RMSE and ubRMSE improvements of 0.027 compared to MODIS albedo products. The high-resolution data was then used to identify firn/snow extents using a 0.37 threshold, facilitating the extraction of long-term firn-line altitudes (FLA) to indicate the glacier dynamics. Our findings revealed that a consistent decline in summer albedo across the Karakoram over the past three decades, signifying a darkening of glacier surfaces that increased solar radiation absorption and intensified melting. The reduction in albedo showed spatial heterogeneity, with slower reductions in the western and central Karakoram (-0.0005-0.0005 yr(-1)) compared to the eastern Karakoram (-0.006 similar to -0.01 yr(-1)). Notably, surge- or advance-type glaciers, avalanche-fed glaciers and debris-covered glaciers exhibited slower albedo reduction rates, which decreased further with increasing glacier size. Additionally, albedo reduction accelerated with altitude, peaking near the equilibrium-line altitude. Fluctuations in the albedo-derived FLAs suggest a transition in the dynamics of Karakoram glaciers from anomalous behavior to retreat. Most glaciers exhibited anomalous behavior from 1995 to 2010, peaking in 2003, but they have shown signs of retreat since the 2010s, marking the end of the Karakoram anomaly. These insights deepen our understanding of the Karakoram anomaly and provide a theoretical basis for assessing the effect of glacier anomaly to retreat dynamics on the water resources and adaptation strategies for the Indus and Tarim Rivers.

In potato breeding, maturity class (MC) is a crucial selection criterion because this is a critical aspect of commercial potato production. Currently, the classification of potato genotypes into MCs is done visually, which is time- and labor-consuming. The objective of this research was to use vegetation indices (VIs) derived from unmanned aerial vehicle (UAV) imagery to remotely assign MCs to potato plants grown in trials, representing three different early stages within a multi-year breeding program. The relationships between VIs (GOSAVI - Green Optimized Soil Adjusted Vegetation Index, MCARI2 - Modified Chlorophyll Absorption Index-Improved, NDRE - Normalized Difference Red Edge, NDVI - Normalized Difference Vegetation Index, and OSAVI - Optimized Soil Adjusted Vegetation Index and WDVI - Weighted Difference Vegetation Index) and visual potato canopy status were determined. Further, this study aimed to identify factors that could improve the accuracy (decrease Mean Absolute Error - MAE) of potato MC estimation remotely. Results show that VIs derived from UAV imagery can be effectively used to remotely assign MCs to potato breeding lines, with higher accuracy for the potato B-clones (20 plants per plot) than the A-clones (6 plants per plot). Among the tested VIs, the NDRE allowed for potato MC evaluation with the lowest MAE. Applying NDRE for remote MC estimation using a validation dataset of potato B-clones (100 plants per plot), resulted in an MC estimate with a 0.81 MAE. However, the accuracy of potato MC estimation using UAV image-based methods should be improved by reducing the potato canopy's variability (increasing uniformity) within the plot. This could be achieved by minimizing 1) potato vines bending over the neighboring row, causing vine overlap between plots, and 2) plants damaged by tractor wheels during field operations. En el mejoramiento de la papa, la clase de madurez (CM) es un criterio de selecci & oacute;n crucial porque este es un aspecto cr & iacute;tico de la producci & oacute;n comercial de papa. Actualmente, la clasificaci & oacute;n de los genotipos de papa en MC se realiza visualmente, lo que requiere mucho tiempo y trabajo. El objetivo de esta investigaci & oacute;n fue utilizar & iacute;ndices de vegetaci & oacute;n (VIs) derivados de im & aacute;genes de veh & iacute;culos a & eacute;reos no tripulados (UAV) para asignar de forma remota MCs a plantas de papa cultivadas en ensayos, representando tres etapas tempranas diferentes dentro de un programa de mejoramiento de varios a & ntilde;os. Se determinaron las relaciones entre los VIs (GOSAVI - & Iacute;ndice de Vegetaci & oacute;n Ajustado al Suelo Optimizado Verde, MCARI2 - & Iacute;ndice de Absorci & oacute;n de Clorofila Modificado-Mejorado, NDRE - Borde Rojo de Diferencia Normalizada, NDVI - & Iacute;ndice de Vegetaci & oacute;n de Diferencia Normalizada, y OSAVI - & Iacute;ndice de Vegetaci & oacute;n Ajustado al Suelo Optimizado y WDVI - & Iacute;ndice de Vegetaci & oacute;n de Diferencia Ponderada) y la visualizaci & oacute;n del dosel de la papa. Adem & aacute;s, este estudio tuvo como objetivo identificar factores que podr & iacute;an mejorar la precisi & oacute;n (disminuir el Error Absoluto Medio - MAE) de la estimaci & oacute;n de MC de papa de forma remota. Los resultados muestran que los VI derivados de las im & aacute;genes de UAV se pueden utilizar de manera efectiva para asignar MC de forma remota a las l & iacute;neas de mejoramiento de papa, con mayor precisi & oacute;n para los clones B de papa (20 plantas por parcela) que para los clones A (6 plantas por parcela). Entre los VI probados, el NDRE permiti & oacute; la evaluaci & oacute;n de la MC de papa con el MAE m & aacute;s bajo. La aplicaci & oacute;n de NDRE para la estimaci & oacute;n remota de MC utilizando un conjunto de datos de validaci & oacute;n de clones B de papa (100 plantas por parcela), result & oacute; en una estimaci & oacute;n de MC con un MAE de 0.81. Sin embargo, la precisi & oacute;n de la estimaci & oacute;n de la MC de la papa utilizando m & eacute;todos basados en im & aacute;genes UAV debe mejorarse reduciendo la variabilidad del dosel de la papa (aumentando la uniformidad) dentro de la parcela. Esto podr & iacute;a lograrse minimizando 1) los tallos de papa que se doblan sobre el surco vecino, lo que causa la superposici & oacute;n de follaje entre las parcelas, y 2) las plantas da & ntilde;adas por las ruedas de los tractores durante las operaciones de campo.

In recent decades, increases in severe drought, heat extremes, and pest burden have contributed to increased global tree mortality. These risks are expected to be exacerbated under projected climate change. So far, observations of tree mortality are mainly based on manual field surveys with limited spatial coverage. The lack of accurate tree mortality data over large areas has limited the development and applications of tree mortality models. However, a combination of high-resolution remote sensing data, such as aerial imagery and automated imagery analysis, may provide a solution to this problem. In this study, we analysed the dynamics and drivers of forest canopy mortality in 117 366 ha of boreal forest in Southeast Finland, between 2017 and 2023. For this purpose, we first developed a fully convolutional semantic segmentation model to automatically segment forest canopy mortality from aerial imagery in 2017, 2020, and 2023 with a spatial resolution of 0.5 m. Secondly, we trained the model using a dataset consisting of 32555 canopy mortality segments manually delineated from aerial imagery from various geographic regions in Finland. The trained model showed high accuracy in detecting forest canopy mortality (with an F1 score of 0.86-0.93) when tested using an independent test set. To estimate standing deadwood volume, we combined the observed yearly forest canopy mortality with open forest resource information based on extensive field campaigns and airborne laser scanning. In our study area, forest canopy mortality increased from 23.4 ha (0.02 % of the study area) to 207.8 ha (0.18 %) between 2017 and 2023. Consequently, standing deadwood volume was estimated to increase from 5192 m3 (0.04 m3/ha) to 52800 m3 (0.45 m3/ha) during the study period. Both the volume of standing deadwood and the extent of forest canopy mortality increased exponentially. The majority of the forest canopy mortality occurred in Norway sprucedominated forests (64.1-77.3 %) on relatively fertile soils (81.6-84.7 %) while 20-25 % of the forest canopy mortality occurred in Scots pine-dominated forests. The average age of stands where mortality was observed was between 60 and 70 years old (2017 = 69.7 years and 2023 = 62.6 years), indicating that mature forests were more susceptible to mortality than younger stands. Our findings highlight an exponential increase in forest canopy mortality over a relatively short time span (6 years). The increasing risk of tree mortality in boreal forests underlines the urgent need for large-scale and spatially accurate monitoring to keep up to date with fast-paced changes in boreal forest mortality. As climate change increases drought, extreme heat and bark beetle outbreaks, consistent canopy mortality mapping is essential for implementing timely risk management measures in forestry.

This study presents a deep learning model created for enabling comprehensive wildfire control by seamlessly combining satellite images, weather data and terrain details. Current systems face challenges in comprehensively analyzing these factors due to limitations in data integration, dynamic fire behavior prediction, and post-fire ecological impact evaluation. By improving detection and accurate assessment of impact, the system addresses all aspects of wildfire management from forecasting to post event analysis. The model integrates soil quality examination and vegetation regrowth simulation Using image analysis and state of the art deep learning methods. This holistic approach of Image analysis employs Convolutional Neural Networks (CNN) for predicting wildfire risk and Recurrent Neural Networks (RNN) for assessing soil and hydrological effects. This adaptable approach, which aims to transform the way fire control is done, can be readily adjusted to changing conditions and takes correlations between different aspects into account. It surpasses conventional techniques by including soil quality analysis, vegetation regrowth modeling, and vegetation damage evaluation. The adaptable nature of this method proves invaluable, in lessening the impact of wildfires with a focus, on evaluating vegetation damage and promoting restoration.

Worldwide, forests have been harassed by fire in recent years. Either by human intervention or other reasons, the history of the burned area is increasing considerably, harming fauna and flora. It is essential to detect an early ignition for fire-fighting authorities can act quickly, decreasing the impact of forest damage impacts. The proposed system aims to improve nature monitoring and improve the existing surveillance systems through satellite image recognition. The soil recognition via satellite images can determine the sensor modules' best position and provide crucial input information for artificial intelligence-based systems. For this, satellite images from the Sentinel-2 program are used to generate forest density maps as updated as possible. Four classification algorithms make the Tree Cover Density (TCD) map, consisting of the Gaussian Mixture Model (GMM), Random Forest (RF), Support Vector Machine (SVM), and K-Nearest Neighbors (K-NN), which identify zones by training known regions. The results demonstrate a comparison between the algorithms through their performance in recognizing the forest, grass, pavement, and water areas by Sentinel-2 images.