Investigating the performance of super-resolved remote sensing images on coastline segmentation with deep learning based methods DOI Open Access
İlhan Pala, Uğur Algancı

International Journal of Engineering and Geosciences, Год журнала: 2024, Номер 10(1), С. 93 - 106

Опубликована: Окт. 11, 2024

The use of satellite imagery in critical areas, such as environmental monitoring and natural disaster management, is becoming increasingly important. Applications like coastal detecting erosion, tracking land changes demand high accuracy detailed analysis. Traditional methods for coastline segmentation are often limited by the low resolution (LR) complexity imagery. To address this challenge, Super Resolution (SR) algorithms employed to enhance images, which particularly beneficial when examining areas with intricate structures, coastlines. In context, integration SR techniques presents an innovative approach achieving greater efficiency image study, images was enhanced using Generative Adversarial Networks (SRGAN) model. Thanks flexible architecture SRGAN model, it successfully adapted work yielding satisfactory results. Coastal performed low-resolution, super-resolved, high-resolution Gokturk-1 (GT-1) employing U-net, LinkNet, DeepLabV3+ models comparison. results indicated that increment significantly affects success. Additionally, better performance achieved U-net LinkNet models. Although model effective segmentation, tends capture less detail compared other two Overall, combination produced were closest reality

Язык: Английский

GIS-Based Analysis of Land Surface Characteristics and Urban Heat Islands in Metropolitan Cities of India DOI Open Access
Rupesh Kumar Gupta

International Journal of Engineering and Geosciences, Год журнала: 2025, Номер 10(3), С. 440 - 455

Опубликована: Апрель 18, 2025

This study compares the Urban Heat Island (UHI) phenomenon in two Indian metropolitan cities: Lucknow and Delhi. helps to understand comparative impact of urbanisation on UHI effect mid-sized large urban cities using multi-temporal satellite data index-based Analysis. MODIS were used examine UHI, while Landsat 8 helped extract land surface features both cities. Various indices, such as NDVI, MBI, MNDWI, NDBaI, utilised characteristics area GIS-based tools methods. The findings indicate that about 18.15% is classified a High Potential Zone, compared approximately 17.17% Land temperatures (LST) rose from 38.11°C 30.41°C 2000 46.17°C 39.15°C 2023. Similarly, Delhi, LST values increased 38.35°C 24.49°C 47.27°C 32.93°C These zones are typically found locations with high built-up density unplanned development activities. identifies negative correlation between presence green blue spaces, which can help reduce intensity UHI. research emphasises importance understanding managing highly urbanised areas, this knowledge will assist policymakers stakeholders enhancing livability sustainability within

Язык: Английский

Процитировано

1
A five-year Study Using Sentinel-5P Data Observing Seasonal Dynamics and Long-term Trends of Atmospheric Pollutants DOI Open Access

Hasan Mohamed,

Abubakr Hassan,

Abdelrahim Elhag

и другие.

International Journal of Engineering and Geosciences, Год журнала: 2025, Номер 10(2), С. 262 - 271

Опубликована: Янв. 25, 2025

Air pollution is an escalating concern for both environmental sustainability and public health, exacerbated by urbanization industrial growth. In Saudi Arabia, pollutants primarily from activities vehicle emissions present significant health hazards. This study utilizes data the Sentinel-5P satellite to analyze variations in Carbon Monoxide (CO), Nitrogen Dioxide (NO₂), Sulfur (SO₂), Particulate Matter (PM2.5) over a five-year period, January 2019 December 2023. The was processed using Google Earth Engine (GEE) produce monthly seasonal averages, while ArcGIS Pro used map trends spatial distribution. results reveal distinct fluctuations levels, with CO peaking between March-May July-September but showing overall decline. NO₂ SO₂ exhibit highs slight upward trends, likely linked output traffic emissions. PM2.5, most harmful pollutant human consistently surpasses World Health Organization (WHO) limits, especially during high-emission periods. These findings underscore urgency of adopting targeted measures mitigate critical times safeguard health. spikes, particularly densely populated regions, highlight need improved policies technologies effectively monitor manage air quality

Язык: Английский

Процитировано

0
Climate patterns in Europe: A focus on ten countries through remote sensing DOI Open Access
Volkan Yılmaz

International Journal of Engineering and Geosciences, Год журнала: 2025, Номер 10(3), С. 398 - 418

Опубликована: Март 20, 2025

Leveraging high-temporal resolution remote sensing data enables the investigation of impacts climate change with unprecedented detail and accuracy. This approach provides consistent observations, allowing for tracking short-term fluctuations long-term trends in patterns. The majority existing studies focus on local impacts, overlooking broader national-scale implications. research addresses this gap, examining effects European countries, i.e., Türkiye, Germany, Belgium, United Kingdom (UK), France, Spain, Switzerland, Italy, Ukraine Poland from 2001 to 2023, emphasizing interconnected nature need comprehensive strategies a national scale. involved examination essential environmental variables, such as precipitation (PCP), land surface temperature (LST), evapotranspiration (ET), potential (PET), normalized difference vegetation index (NDVI), condition (VCI), (TCI), health (VHI) forest area loss (FAL) through an extensive time-series analysis. primary aim was reveal temporal patterns within these datasets. Subsequently, pair-wise correlations among datasets were computed, offering valuable insights into complex interconnections factors used. experiments revealed that UK experienced significant decline PCP, while exhibited higher rates LST increase. France Italy showed ET rates; highest rate PET had more pronounced rise health. study found strong positive (average 0.72) between PET. Additionally, notable correlation NDVI 0.55) VCI 0.42). PCP generally negative other correlated both 0.56). is expected contribute understanding

Язык: Английский

Процитировано

0
Advanced stepwise machine learning integration of near-real-time precipitation products in China's flood-prone basins DOI
Lingxue Liu,

Huajin Lei

Atmospheric Research, Год журнала: 2025, Номер unknown, С. 108197 - 108197

Опубликована: Май 1, 2025

Язык: Английский

Процитировано

0
Investigating the performance of super-resolved remote sensing images on coastline segmentation with deep learning based methods DOI Open Access
İlhan Pala, Uğur Algancı

International Journal of Engineering and Geosciences, Год журнала: 2024, Номер 10(1), С. 93 - 106

Опубликована: Окт. 11, 2024

The use of satellite imagery in critical areas, such as environmental monitoring and natural disaster management, is becoming increasingly important. Applications like coastal detecting erosion, tracking land changes demand high accuracy detailed analysis. Traditional methods for coastline segmentation are often limited by the low resolution (LR) complexity imagery. To address this challenge, Super Resolution (SR) algorithms employed to enhance images, which particularly beneficial when examining areas with intricate structures, coastlines. In context, integration SR techniques presents an innovative approach achieving greater efficiency image study, images was enhanced using Generative Adversarial Networks (SRGAN) model. Thanks flexible architecture SRGAN model, it successfully adapted work yielding satisfactory results. Coastal performed low-resolution, super-resolved, high-resolution Gokturk-1 (GT-1) employing U-net, LinkNet, DeepLabV3+ models comparison. results indicated that increment significantly affects success. Additionally, better performance achieved U-net LinkNet models. Although model effective segmentation, tends capture less detail compared other two Overall, combination produced were closest reality

Язык: Английский

Процитировано

2