Natural disaster impact assessment is of the utmost significance for post-disaster recovery, environmental protection, and hazard mitigation plans. With their recent usage in landslide susceptibility mapping, deep learning (DL) architectures have proven their efficiency in many scientific studies. However, some restrictions, including insufficient model variance and limited generalization capabilities, have been reported in the literature. To overcome these restrictions, ensembling DL models has often been preferred as a practical solution. In this study, an ensemble DL architecture, based on shared blocks, was proposed to improve the prediction capability of individual DL models. For this purpose, three DL models, namely Convolutional Neural Network (CNN), Recurrent Neural Network (RNN), and Long Short-Term Memory (LSTM), together with their ensemble form (CNN–RNN–LSTM) were utilized to model landslide susceptibility in Trabzon province, Turkey. The proposed DL architecture produced the highest modeling performance of 0.93, followed by CNN (0.92), RNN (0.91), and LSTM (0.86). Findings proved that the proposed model excelled the performance of the DL models by up to 7% in terms of overall accuracy, which was also confirmed by the Wilcoxon signed-rank test. The area under curve analysis also showed a significant improvement (~4%) in susceptibility map accuracy by the proposed strategy.

Annually, a considerable quantity of forest is burned on a global scale. Therefore, it is essential to obtain precise and fast information regarding the size of burned regions in order to effectively monitor the adverse consequences of wildfires. The objective of this investigation is to indicate the effectiveness and usefulness of a deep learning (DL) architecture, such as Convolutional Neural Networks (CNNs), in the mapping of areas affected by fire, employing an eXplainable artificial intelligence (XAI) algorithm known as SHapley Additive exPlanations (SHAP) with accuracy evaluation criteria. Furthermore, this paper presents the evaluation of the Çanakkale-Kizilkeçili village wildfire. The research investigated the impacts of a variety of spectral indices, including the normalized burn index (NBR), differentiated normalized burn index (dNBR), Green-Red Vegetation Index (GRVI), simple ratio vegetation index (RVI), and normalized difference vegetation index (NDVI). At the end of the training process, the model achieved a training accuracy of approximately 0.99, with model loss values converging to approximately 0.1. The findings of the burned area identification analysis indicate that by incorporating spectral indices as supplementary information, the CNN model achieved a high level of accuracy, with an overall accuracy of 98.88% and a Kappa Coefficient of 0.98. Additionally, the SHAP technique was employed to gain insights into the output of the models. The feature importances of the spectral bands were determined through the SHAP analysis of the CNN model. Hence, the significance of the auxiliary data generated by the NBR, dNBR, and NDVI indices was identified as being the highest among the original bands and auxiliary data employed in this investigation.

Accurate extraction of building features from remotely sensed data is essential for supporting research and applications in urban planning, land management, transportation infrastructure development, and disaster monitoring. Despite the prominence of deep learning as the state-of-the-art (SOTA) methodology for building extraction, substantial challenges remain, largely stemming from the diversity of building structures and the complexity of background features. To mitigate these issues, this study introduces DeepSwinLite, a lightweight architecture based on the Swin Transformer, designed to extract building footprints from very high-resolution (VHR) imagery. The model integrates a novel local-global attention module to enhance the interpretation of objects across varying spatial resolutions and facilitate effective information exchange between different feature abstraction levels. It comprises three modules: multi-scale feature aggregation (MSFA), improving recognition across varying object sizes; multi-level feature pyramid (MLFP), fusing detailed and semantic features; and AuxHead, providing auxiliary supervision to stabilize and enhance learning. Experimental evaluations on the Massachusetts and WHU Building Datasets reveal the superior performance of DeepSwinLite architecture when compared to existing SOTA models. On the Massachusetts dataset, the model attained an OA of 92.54% and an IoU of 77.94%, while on the WHU dataset, it achieved an OA of 98.32% and an IoU of 92.02%. Following the correction of errors identified in the Massachusetts ground truth and iterative enhancement, the model’s performance further improved, reaching 94.63% OA and 79.86% IoU. A key advantage of the DeepSwinLite model is its computational efficiency, requiring fewer floating-point operations (FLOPs) and parameters compared to other SOTA models. This efficiency makes the model particularly suitable for deployment in mobile and resource-constrained systems.

1. Forests play a critical role in ecosystem services that control climate and rainfall patterns, prevent erosion, and provide a home for plants and animals. Globally, deforestation is a significan...

Son yıllarda hızlı gelişen türler ve klonların belirli üretim teknik ve sistemlerle yetiştirilmesi, odun hammadde arzının karşılanması noktasında büyük öneme sahiptir. Bu bağlamda, dünyada ve Türkiye’de kavak türleri ve klonları kullanılarak geniş ölçekte endüstriyel plantasyonlar oluşturulmaktadır. Türkiye’de yetiştirilen kavak türlerinin ayrımı, yayılış alanlarının tespiti, izlenmesi ve değerlendirilmesi geleceğe yönelik odun hammaddesi yönetim ve planlamalarına altlık oluşturmaktadır. Son yıllarda, orman alanlarının izlenmesi için tercih edilen ve yaygın kullanılan yaklaşım, uzaktan algılama ve coğrafi bilgi sistemlerinin entegrasyonudur. Kavak yetişme sahalarının diğer türlerin yetişme alanlarından ayrılması ve alansal tespitinin yapılabilmesi için türlerin spektral özelliklerinin bilinmesi gerekir. Enerji nesne arasındaki ilişki, nesnelerin spektral imzası olarak tanımlanmakta ve görüntü üzerinden farklı spektral özelliklere sahip nesnelerin ayırt edilmesinde kullanılan temel veri kaynağı niteliğindedir. Bu çalışma kapsamında, farklı endüstriyel kavak tür ve klonlarını içerisinde barındıran Sakarya-Akyazı, Uşak-Sivaslı, Afyonkarahisar-Çay ve Kırşehir-Özbağ ilçelerinde belirlenen örnekleme alanlarında, Populus deltoides Bartr. “Samsun (I-77/51)”, Populus x euramericana “I-214”, Populus x euramericana “I-45/51”, Populus nigra L. “Kocabey (TR-77/10)” klonları ve Populus nigra L. (Karakavak) türü için spektroradyometre cihazı yardımıyla spektral ölçüm çalışmaları gerçekleştirilmiştir. Kavak tür ve klonlarına ait, 380-720 nm dalga boyu aralığının mavi bandında tüm kavak tür ve klonlarının benzer spektral karakteristiğe sahip olduğu fakat yeşil ve kırmızı/kırmızı-kenar bantlarda spektral yansıma değerlerinde farklılıklar gösterdiği tespit edilmiştir. Elektromanyetik spektrumun özellikle kızılötesi bölgesinde Populus nigra L. “Kocabey (TR 77/10)” klonunun diğerlerinden daha rahat ayırt edilebildiği gözlenmiştir. Elektromanyetik spektrumun kısa dalga kızılötesi bölgesinde ise klonal spektral özelliklerinin diğerlerinden farklılaştığı, bu sayede görünür veya yakın kızılötesi bölgede ayırt edilemeyen tür ve klonların bu bölgede ayırt edilebildiği tespit edilmiştir.

Object-Based Image Analysis (OBIA) is a method employed in the field of remote sensing with the objective of enhancing classification accuracy. This is achieved by focusing on image segments comprising groups of pixels, rather than evaluating individual pixels. By addressing the limitations of traditional pixel-based methods, OBIA is employed for the classification of segments based on their attributes. The present study evaluates the use of OBIA-based classification in conjunction with deep learning and machine learning classifiers. A study area, approximately 210 km² located in Ankara, was selected and SPOT-6 imagery with a spatial resolution of 1.5 meters and 4 spectral bands (red, green, blue and near infrared) was employed for this purpose. In the segmentation stage, a multiresolution segmentation approach was employed, and classification process was conducted using a Convolutional Neural Network (CNN) and Extreme Gradient Boosting (XGBoost). The CNN classifier demonstrated superior performance compared to the XGBoost algorithm, with an improvement of 2.7%. The Shapley Additive Explanations (SHAP) technique, an effective Explainable Artificial Intelligence (XAI) method, was employed to assess the explainability of the classifiers. The SHAP analysis indicated that the HSI transform was the most influential factor in the XGBoost algorithm’s decision-making process whereas the average DN values of the green band were the most effective feature for the CNN model. Global SHAP analyses elucidated the overarching model decision-making process, whereas class-specific analyses furnished insights into the classification of each land use and land cover (LULC) class.