Article
Environmental Sciences
V. S. Martins, D. P. Roy, H. Huang, L. Boschetti, H. K. Zhang, L. Yan
Summary: High spatial resolution commercial satellite data provide new opportunities for terrestrial monitoring. In particular, the near-daily 3 m observations provided by the PlanetScope constellation enable mapping of small and spatially fragmented burns that are not detectable at coarser spatial resolution. This study demonstrates the potential for automated PlanetScope 3 m burned area mapping using a deep learning algorithm called U-Net. The algorithm incorporates spatial and spectral information and was trained with a combination of Landsat-8 and PlanetScope reference data, resulting in high classification accuracy at 3 m resolution.
REMOTE SENSING OF ENVIRONMENT
(2022)
Article
Environmental Sciences
Allan A. Pereira, Renata Libonati, Julia A. Rodrigues, Joana Nogueira, Filippe L. M. Santos, Duarte Oom, Waislan Sanches, Swanni T. Alvarado, Jose M. C. Pereira
Summary: Researchers are working on improving the understanding of fire patterns and changes, and the need for a consistent database about the location and extension of burned areas. They have developed a new algorithm to improve BA mapping accuracy in the Brazilian savannas, which can generate automated products over large areas and long periods.
Article
Environmental Studies
Michail Sismanis, Rizos-Theodoros Chadoulis, Ioannis Manakos, Anastasios Drosou
Summary: The frequency and severity of large, destructive fires have increased recently, causing extensive impacts on the landscape, human population, and ecosystems. This study proposes an unsupervised approach using Sentinel-2 satellite imagery for mapping burned areas, considering the variability of spectral response. It achieves high classification accuracy and can be used as a complementary tool to existing forest fire management services and decision support systems.
Article
Environmental Sciences
Xikun Hu, Yifang Ban, Andrea Nascetti
Summary: The study compares deep learning (DL) models and machine learning (ML) algorithms for mapping burned areas from satellite imagery in three wildfire sites. DL algorithms outperform ML methods in compact burned areas, while ML methods are more suitable for dispersed burn in boreal forests.
Article
Environmental Sciences
Ekhi Roteta, Aitor Bastarrika, Magi Franquesa, Emilio Chuvieco
Summary: This article introduces four burned area tools implemented in Google Earth Engine (GEE) for burned area (BA) mapping using medium spatial resolution sensors. The tools include supervised BA mapping, BA stratified random sampling, and highly accurate BA map verification. Two case studies demonstrate the performance and accuracy of these tools in monitoring wildfires in different regions.
Article
Engineering, Multidisciplinary
Seyd Teymoor Seydi, Mojtaba Sadegh
Summary: Satellite imagery, specifically Landsat, is widely used for mapping and monitoring wildfire burned areas. The new Landsat-9 satellite, with higher radiometric and improved temporal resolution, enables better burned area mapping. A novel deep learning model called convolutional shift-transformer (CST) is proposed, which outperforms other models with an F1-score of over 96% across five large fire case studies globally. CST also reduces computational costs as it only requires a single post-fire image.
Article
Environmental Sciences
Jinxiu Liu, Du Wang, Eduardo Eiji Maeda, Petri K. E. Pellikka, Janne Heiskanen
Summary: Accurate estimation of cropland burned area is essential for air quality modeling and cropland management. Current global burned area products are limited by coarse spatial resolution images, while accurate cropland straw burning identification methods at high temporal and spatial resolution are lacking. This study proposes a novel algorithm using dense Landsat time series and a multi-harmonic model to improve burned area detection accuracy, achieving a superior overall accuracy compared to existing products.
Article
Environmental Sciences
Rob Skakun, Ellen Whitman, John M. Little, Marc-Andre Parisien
Summary: This study compared fire perimeters derived from airborne and satellite imagery with conventional methods, developed prediction models to estimate fire size, and created an adjusted time series of burned areas. By applying statistical adjustments, the study reduced historical overestimations in annual area burned, contributing to a more accurate representation of fire size.
ENVIRONMENTAL RESEARCH LETTERS
(2021)
Article
Geosciences, Multidisciplinary
Jiyu Liu, David Freudenberger, Samsung Lim
Summary: This study utilized a multi-resolution segmentation method and a hierarchical classification framework based on expert knowledge to classify burned areas and land-uses in Kangaroo Island, Australia. The results demonstrated that the object-based image classification framework, using multi-source data, achieved higher accuracy in classification and can contribute to improved fire management and control.
GEOMATICS NATURAL HAZARDS & RISK
(2022)
Article
Environmental Sciences
Rogerio G. Negri, Andrea E. O. Luz, Alejandro C. Frery, Wallace Casaca
Summary: This paper introduces a unified data-driven framework using multispectral images and statistical modeling to map areas damaged by forest fires. Through case studies with remote-sensing images, the method outperforms other evaluated methods in terms of accuracy, F1 score, and kappa coefficient. The proposed method provides spatial-adherence mappings that match estimates reported by the MODIS Burn Area product.
Article
Geography, Physical
Mingjun He, Junyu He, Yajun Zhou, Liyuan Sun, Shuangyan He, Cong Liu, Yanzhen Gu, Peiliang Li
Summary: This study utilized remote sensing technology to observe and classify the coral reefs in the Xisha Islands, providing important information for the management and conservation of the coral reef ecosystems.
GISCIENCE & REMOTE SENSING
(2023)
Article
Remote Sensing
Alana K. Neves, Manuel L. Campagnolo, Joao M. N. Silva, Jose M. C. Pereira
Summary: Portugal is the most fire-prone country in Southern Europe. This study used time series of satellite imagery to improve fire mapping and analyze the intra-annual fire variability in Portugal. The Portuguese Annual Fire Atlas burned area patches were disaggregated into individual events based on their date of occurrence estimated from Landsat temporal series. The resulting Monthly Fire Atlas achieved high accuracy in identifying fire dates.
INTERNATIONAL JOURNAL OF APPLIED EARTH OBSERVATION AND GEOINFORMATION
(2023)
Article
Environmental Sciences
Rob Skakun, Guillermo Castilla, Juha Metsaranta, Ellen Whitman, Sebastien Rodrigue, John Little, Kathleen Groenewegen, Matthew Coyle
Summary: Wildfires are a significant issue in Canada, and their frequency is expected to increase due to climate change. To accurately analyze trends in burned area and understand the impacts of fire frequency, duration, and extent, a long-term and reliable dataset is needed. In this study, the National Burned Area Composite (NBAC) dataset was extended to include data from 1986 to 2020. The dataset consists of annual maps in polygon format, with different mapping methods and data sources used to delineate the burned area. The results show that the majority of the burned area was derived from change detection methods using Landsat satellite imagery. Confidence intervals were calculated for each year to reflect the accuracy and contribution of different data sources, and the NBAC dataset had narrower confidence intervals compared to the Canadian National Fire Database (CNFDB). Furthermore, the NBAC dataset identified additional fire events that were missing in the CNFDB, highlighting its importance for regional fire analysis and ecological studies.
Article
Environmental Sciences
Miguel A. Belenguer-Plomer, Mihai A. Tanase, Emilio Chuvieco, Francesca Bovolo
Summary: This study analyzed the use of convolutional neural networks for mapping burned areas by combining radar and optical datasets. The optimal CNN settings and sensor integration were determined based on land cover class and data type. Increasing network complexity did not improve accuracy in burned area mapping.
REMOTE SENSING OF ENVIRONMENT
(2021)
Article
Environmental Sciences
Mingyue Wei, Zhaoming Zhang, Tengfei Long, Guojin He, Guizhou Wang
Summary: This study generated a high-resolution global burned area product based on remote sensing technology and analyzed the changes in global BA from 2015 to 2019. It was found that the total area of global BA was relatively stable during this period, but significant differences existed among continents and regions, particularly in the Amazon and Australia.
Article
Engineering, Geological
Emre Senkal, Gordana Kaplan, Ugur Avdan
Summary: This study investigated the accuracy of UAV products in an archaeological area, showing significant differences between UAV images and those collected using conventional methods. The use of control points improved the results significantly.
INTERNATIONAL JOURNAL OF ENGINEERING AND GEOSCIENCES
(2021)
Article
Environmental Sciences
Dilek Kucuk Matci, Resul Comert, Ugur Avdan
Summary: This study determined the rate of urbanization and industrialization in Eskisehir city center in Turkey between 1984 and 2020, and simulated the area needed for urban and industrial areas in 2030 using a hybrid model. The analysis of map changes and simulation predictions showed that the growth of urban and industrial areas will result in a decrease in agriculture and other natural areas.
JOURNAL OF THE INDIAN SOCIETY OF REMOTE SENSING
(2022)
Article
Environmental Sciences
Firat Erdem, Rutkay Atun, Zehra Yigit Avdan, Ilknur Atila, Ugur Avdan
Summary: Drought is defined as a situation where the amount of precipitation in a region is less than the amount of evaporation. Human activities such as population growth, industrialization, deforestation, and excessive irrigation in agriculture are major factors contributing to drought. This study used remote sensing and GIS methods to analyze the drought change in the Van Lake Basin from 1989 to 2019, showing an increasing trend in drought severity over the years.
EGYPTIAN JOURNAL OF REMOTE SENSING AND SPACE SCIENCES
(2021)
Article
Engineering, Geological
Nese Basaran, Dilek Kucuk Matci, Ugur Avdan
Summary: Losses of forest area have significant negative impacts on ecosystems, economies, and societies. Monitoring and analyzing this process is crucial for minimizing these negative effects and promoting urban development. This study used the Google Earth Engine to examine the factors causing forest losses in the Mediterranean Region and considered various factors influencing forest area changes.
INTERNATIONAL JOURNAL OF ENGINEERING AND GEOSCIENCES
(2022)
Article
Environmental Sciences
Sayed Ishaq Deliry, Emrah Pekkan, Ugur Avdan
Summary: Satellite remote sensing products are increasingly important in water resources management. This study estimated the water budget components of the Kizilirmak River Basin using satellite observations and models, and found good consistency in precipitation data but large uncertainties in evapotranspiration and terrestrial water storage. Inferred runoff from remote sensing and model outputs showed differences from observed streamflow measurements, with the model demonstrating better consistency. This study revealed the strengths and limitations of satellite remote sensing and models in estimating water budget.
JOURNAL OF THE INDIAN SOCIETY OF REMOTE SENSING
(2022)
Article
Computer Science, Interdisciplinary Applications
Dilek Kucuk Matci, Ugur Avdan
Summary: This study proposes a method for automatically labeling images without a training phase. By using bands in the image and Corine data, a database is created by examining spectral characteristics of land classes from sample images. The unlabelled classes are evaluated using this database, and the relevant label is assigned. The developed approach is tested in several regions in Turkey and Greece and achieves high accuracy.
EARTH SCIENCE INFORMATICS
(2022)
Article
Environmental Sciences
Dilek Kucuk Matci, Gordana Kaplan, Ugur Avdan
Summary: With increased urbanization, industrialization, and the use of fossil fuels, air pollution has become a pressing global issue. The COVID-19 pandemic led to lockdowns worldwide, prompting new research questions. This study used remote sensing data to analyze air and temperature parameters in different land cover classes in Turkey and investigate their correlation. The findings showed a significant decrease in NO2 in urban areas, which is valuable for long-term strategies to reduce global air pollution. Future research should conduct similar investigations in other locations and evaluate changes in air quality metrics among different classes.
ENVIRONMENTAL MONITORING AND ASSESSMENT
(2022)
Article
Computer Science, Information Systems
Gordana Kaplan, Resul Comert, Onur Kaplan, Dilek Kucuk Matci, Ugur Avdan
Summary: This study successfully extracted the building inventory information using LiDAR derivatives data and machine learning methods, and investigated the important data such as building height and footprint area. The results show that these data can be further utilized in applications such as structural health monitoring.
ISPRS INTERNATIONAL JOURNAL OF GEO-INFORMATION
(2022)
Article
Geography, Physical
Firat Erdem, Nuri Erkin Ocer, Dilek Kucuk Matci, Gordana Kaplan, Ugur Avdan
Summary: Monitoring trees is essential for forest management, urban plant monitoring, vegetation distribution, change monitoring, and establishing sustainable agricultural systems. This study aimed to automatically detect, count, and map apricot trees in orthophotos through deep learning algorithms. The results showed that Mask R-CNN performed better than U-Net in tree detection, counting, and mapping tasks.
PHOTOGRAMMETRIC ENGINEERING AND REMOTE SENSING
(2023)
Article
Geosciences, Multidisciplinary
Ugur Avdan, Gordana Kaplan, Dilek Kucuk Matci, Zehra Yigit Avdan, Firat Erdem, Ece Tugba Mizik, Ilknur Demirtas
Summary: Soil salinity, as a significant environmental problem, requires timely and accurate mapping and monitoring. This study investigates and compares soil salinity models developed from remote sensing data with different spectral and spatial resolutions. The results show that higher spatial resolution leads to better model prediction, but also increases the complexity of model development.
PHYSICS AND CHEMISTRY OF THE EARTH
(2022)
Article
Environmental Sciences
Nalan Demircioglu Yildiz, Firat Erdem, Seyma Berk Acet, Ugur Avdan
Summary: The increase in population in cities leads to economic, social, and environmental problems and the resulting microclimatic conditions can cause more environmental issues. Urban geometry is one of the factors affecting urban climate. Factors such as sky view factor (SVF), building view factor (BVF), and tree view factor (TVF) are used to determine urban geometry, but their effects on the thermal state of the urban environment are not well-studied. Understanding the relationship between urban growth, land surface changes, and thermal conditions is crucial for sustainable urban planning.
ENVIRONMENTAL SCIENCE AND POLLUTION RESEARCH
(2023)
Article
Remote Sensing
Ugur Avdan, Dilek Kucuk Matci, Gordana Kaplan, Zehra Yigit Avdan, Firat Erdem, Ilknur Demirtas, Ece Tugba Mizik
Summary: This study evaluates the impact of atmospheric correction on soil salinity determination using Landsat 8 and Sentinel-2 data. The results show that atmospheric correction significantly affects the relationship between spectral indices and in situ salinity measurement values.
Article
Agriculture, Multidisciplinary
Murat Kuruca, Dilek Kocuk Matci, Ugur Avdan
Summary: The results of the study demonstrate that Gokturk-2 satellite images are effective for mapping burnt forest areas. Support vector machine classification shows similar accuracy and overall accuracy for Gokturk-2 and Landsat-8 images, while Worldview-2 performs better in general accuracy.
TURKISH JOURNAL OF AGRICULTURE AND FORESTRY
(2021)