Egyptian Journal of Remote Sensing and Space Sciences最新文献

Accurate geoid modeling is significant in geodetic, geological, and environmental sciences. Owing to challenges in establishing reference stations, particularly in rugged terrains, such as in Northern Vietnam, leveraging global geopotential models (GGMs) is imperative. Herein, we proposed a superior method that integrates GGMs with advanced artificial intelligence (AI) algorithms to enhance the accuracy and spatial resolution of regional geoid models. A total of six contemporary GGMs (XGM2019e_2159, SGG-UGM-2, SGG-UGM-1, GECO, EIGEN-6C4, and EGM2008) were systematically evaluated to identify the optimal GGM that represents the Earth’s gravitational field in Northern Vietnam. Subsequently, sophisticated AI algorithms, including tree-based ensembles, support vector machines, Gaussian linear regression, regression trees, and linear regression models, were implemented. These AI algorithms were trained on the integrated global navigation satellite system (GNSS) leveling data and corresponding height anomalies to capture complex relationships in the geopotential field. Among the six investigated GGMs, XGM2019e_2159 shows optimal performance for Northern Vietnam, displaying a standard deviation of ±0.17 m. Rigorous assessment results from cross-validation and validation against independent datasets demonstrate satisfactory accuracy across all considered models. However, the Gaussian process regression model with an exponential kernel exhibits marginal superiority, boasting a standard deviation of approximately 0.07 m. This model is therefore chosen for the construction of the geoid model by integrating ground data with optimal GGMs, which shows superior performance, particularly in challenging topographic and geophysical conditions, thereby contributing to a marked improvement in the realized spatial resolution.

Despite comprising only one per cent of global forests, mangroves provide vital ecological and economic benefits to their ecosystems. Due to its decreasing extent over the past decade, there is a rise in research innovations supporting mangrove conservation. Specifically, consumer-grade Unmanned Aerial Vehicles (UAV) were proven effective as potential remote sensing alternatives to support mangrove research and monitoring in recent studies. As most studies use custom UAV-mounted sensors for mangrove species classification, similar studies using a UAV’s default red–green–blue (RGB) cameras were scarce. This study explores the potential of high-resolution RGB aerial images through state-of-the-art object detection algorithm, YOLOv5 to detect the dominant Rhizophora mangroves in Sarawak, Malaysia. A total of 400 RGB images were equally selected from two study areas and allocated into three datasets, two corresponding to each study area and one combining all images. The annotation process was performed using a previously proposed novel method, assisted by YOLOv5 for a semi-automated annotation process with expert verification. Systematic training experiments were conducted to select an optimal epoch size across models trained with each dataset. The final models produced an average true positive rate of 73.8% and 71.7% for each study site, while the combined dataset model produced an average true positive rate of 73.7%. Overall, this study demonstrated the potential of UAV-based RGB images and deep learning object detection architectures to identify specific mangrove objects, while also highlighting key considerations for similar future research.

The diverse range of uses of unmanned aerial vehicles has garnered significant attention in research. The scientific literature that supports the data obtained from UAVs recording information from various sensors is presented in this manuscript. It summarizes current developments in remote sensing, including radar, photogrammetry, thermal imaging, light detection and ranging sensors (LiDAR), data gathering, and analysis. It is predicated on the instruments’ ability to gather and analyze accurate data. To identify some of the most urgent research problems, it also shows surveys based on research methodologies. The present research focuses on the proliferation and social effects of unmanned aerial vehicles (UAVs). It also encourages novice researchers to pursue this area of study and suggest novel approaches to the design or setup of these flying machines. UAVs have entirely transformed due to advancements in internet technology and current technologies which include camera defects, environmental monitoring, charging, impediments, crop monitoring, energy consumption, military applications, and technology gaps.

There is very little to no literature on the use of geotagging to monitor crops from aerial photos, even though many technologies have been created to do so. Current crop monitoring methods, relying on field data and lab analysis, are inefficient due to high labor, time, and potential harm, limiting their broad use. With the use of vegetation indices (VI) and geotagging, this paper highlights the benefits of crop-specific monitoring with unmanned aerial vehicles (UAV). This study systematically searched the original articles published from the 1st of January 2016 to the 7th of October 2021 in the databases of Scopus, ScienceDirect, Commonwealth Agricultural Bureaux (CAB) Direct, and Web of Science (WoS) using Boolean string: “aerial imagery” AND “vegetation index” OR “vegetation indices“ AND “crop”. Out of the papers identified, 28 eligible studies did meet our inclusion criteria and were evaluated. This review thoroughly discusses the advantages of aerial imagery, vegetation indices, and geotagging tools in the context of crop monitoring. It was found that geotagged crop monitoring using UAV empowers farmers with data-driven insights using vegetation indices, enabling them to make informed decisions before acting, transforming agriculture towards a digital future. This study offers valuable insights for researchers and industry players, helping them identify effective and context-specific crop monitoring strategies for diverse plantations, crops, and budgets. Moreover, by utilizing the advanced computational capabilities of artificial intelligence (AI), we can analyze a wide range of vegetation indices to gain a comprehensive understanding of crop health and conduct accurate predictions.

Land Use Land Cover (LULC) change is a complex phenomenon driven by various natural and anthropogenic factors, significantly impacting carbon storage potential. By applying integrated models of ANN-CA Markov, GeoDetector, and InVEST model, this study aimed to analyze LULC change, their driving factors, and implications on carbon storage in the Forest Management Unit (FMU) of Ampang Plampang in West Nusa Tenggara, Indonesia. Several data sources were utilized in the modelling approach, including DEM (Digital Elevation Model), topographical map, Landsat imageries (2011, 2016, 2021), measured carbon density (above ground, below ground, soil, dead organic), and socio-economic data (number of populations, farmer, and agricultural land). The dryland forest in the study area constitutes the most extensive LULC that has experienced significant declines due to deforestation, predominantly transforming into agricultural land, and these are predicted to continue until 2031 with different magnitudes. The significant driving factors of LULC change were elevation, population pressure on land, and distance from settlement. The LULC change also greatly influenced the decline of carbon storage historically (2011–2016) and in projected LULC (2026–2031). The conversion of forested areas to non-forest LULCs has released carbon emissions of about 1.89 Mt CO₂-eq. The study findings implied that the integration of ANN-CA Markov, GeoDetector, and InVEST models has been helpful for comprehending complicated interactions among LULC change, driving factors, and carbon dynamics. The results also contribute to the scientific knowledge base for land management decision-making and policy formulation. Effective management of LULC changes through low carbon development is suggested to mitigate the loss of carbon storage capacities, foster sustainable development goals (SDGs), support Nationally Determined Contribution (NDC), and improve ecosystem resilience.

The Spatiotemporal variability of air quality is influenced by various factors over time. The objectives of this research are to create prediction models for Carbon monoxide (CO) and Nitrogen dioxide (NO₂) and determine the factors which that most impact CO and NO₂ monthly using Random Forest Prediction. The methodology relies on Random Forest Prediction to predict air quality monthly in 2021, incorporating eight variables land surface temperature (LST), normalized difference built-up index (NDBI), built-up index (BU index), normalized difference vegetation index (NDVI), digital elevation model (DEM), relative humidity (RH), wind speed (WS), and wind direction (WD). The results indicate that RH, elevation, WD, and LST are the most significant factors influencing CO concentrations, representing 33%, 24%, 12%, and 10% respectively at annual level in 2021. Similarly, WD, WS, RH, elevation and LST are the most importance factors impacting NO₂ concentrations, representing 24%, 21%, 18%, 12%, and 10% respectively at an annual level in 2021. Furthermore, NDBI and BU index had the lowest impact in on both CO and NO₂, with BU index showing a slightly higher percentage in NO₂ models compared to CO models. Regarding cross-validation, the MAE values in CO models range from 0.11 to 0.18, and the RMSE values range from 0.14 to 0.23. Additionally, the MAE values in NO₂ models ranges from 3.78 to 7.30, and RMSE values range from 4.93 to 9.23.

Maize, a vital global crop, faces numerous challenges, including outbreaks. This study explores the use of spectral vegetation indices for the early detection of maize diseases in individual leaves based on crop phenology at the vegetative, tasselling, and maturity stages. The research was conducted in rural areas of Giyani in the Limpopo province, South Africa, where smallholder farmers heavily rely on maize production for sustenance. Fungal and viral diseases pose significant threats to maize crops, necessitating precise and timely disease detection methods. Hyperspectral remote sensing, with its ability to capture detailed spectral information, offers a promising solution. The study analysed spectral reflectance data collected from healthy and diseased maize leaves. Various vegetation indices derived from spectral signatures, including the Normalized difference vegetation index (NDVI), Anthocyanin Reflectance Index (ARI), photochemical Reflectance Index (PRI), and Carotenoid Reflectance Index (CRI) were investigated for their ability to show disease-related spectral variations. The results indicated that during the tasselling stage, the spectral differences had minimum absorption in the blue region. However, a distinct shift in spectral reflectance was observed during the vegetative stage with 70 % increase in reflectance. First derivative reflectance analysis revealed peaks at approximately 715 nm and 722 nm, which were useful in the discrimination of the different growth stages. Generalized Linear Models (GLM) with binomial link functions and Akaike Information Criterion (AIC) showed that individual vegetation indices performed equally well. NDVI (P<0.001) and CRI (P<0.000) showed the lowest AIC values across all growth stages, suggesting their potential as effective disease indicators. These findings underscores the significance of employing remote sensing technology and spectral analysis as essential tools in the endeavours to tackle the difficulties encountered by maize growers, especially those operating small-scale farms, and to advance sustainable farming practices and ensure food security.

The selection of an optimal dataset is crucial for successful remote sensing analysis. The PRISMA hyperspectral sensor (with 240 spectral bands) and Landsat OLI-2 (boasting high dynamic resolution) offer robust data for various remote sensing applications, anticipating their increased demand in the coming years. However, despite their potential, we have not identified a rigorous evaluation of both datasets in geological applications utilizing Machine Learning Algorithms. Consequently, we conduct a comprehensive analysis using Random Forest, a widely-recommended machine learning algorithm, and employ K-fold cross-validation (with K = 2, 5, 10) with grid-search hyperparameter tuning for enhanced performance. Toward this aim, diverse image-processing approaches, including Principal Component Analysis (PCA), Minimum Noise Fraction (MNF), and Independent Component Analysis (ICA), were applied to enhance feature selection and extraction. Subsequently, to ensure better performance of the RF algorithm, this study utilized well-distributed points instead of polygons to represent each target, thereby mitigating the effects of spatial autocorrelation. Our results reveal dataset-hyperparameter dependencies, with PRISMA mainly influenced by max_depth and Landsat 9 by max_features. Employing grid-search optimally balances dataset characteristics and data splitting (folds), generating accurate lithological maps across all K values. Notably, a significant hyperparameter shift at K = 10 produces the best lithological maps. Fieldwork and petrographic investigations validate the lithological maps, indicating PRISMA’s slight superiority over Landsat OLI-2. Despite this, given the dataset nature and band count difference, we still advocate Landsat 9 as a potent multispectral input for future applications due to its superior radiometric resolution.

Mangroves are a crucial part of the coastal ecosystem; thus, precise and up-to-date monitoring is essential to guide regional policies and inform conservation strategies. This study investigates the capabilities of semi-automated remote sensing approaches within a Google Earth Engine framework for national-scale mangrove mapping in Thailand. Remote sensing data from 2018—10,000 data points acquired from Sentinel-1, Sentinel-2, and the Shuttle Radar Topography Mission (SRTM)—was used to train several machine learning models. The Gradient Tree Boost (GTB) proved to be the most reliable, with the least variation in validity (the lowest IQR) and the highest average Overall Accuracy of 96.75 ± 0.63 % compared to the others—96.64 ± 0.72 % for Random Forest (RF); 96.12 ± 0.80 %for Classification and Regression Trees (CART); and 95.43 ± 0.74 % for Support Vector Machines (SVM). Thus, the GTB was instrumental in mapping mangrove distribution with 10-m spatial resolution across Thailand from 2016 to 2022, the period in which the mangrove areas increased by 11 %, reflecting successful conservation efforts over the past decade. The developed framework establishes the foundation for semi-automated mangrove mapping that can be developed for other geographical contexts.