A Global Swin-Unet Sentinel-2 Surface Reflectance-Based Cloud and Cloud Shadow Detection Algorithm for the Nasa Harmonized Landsat Sentinel-2 (Hls) Dataset DOI
Haiyan Huang, David P. Roy, Hugo De Lemos

и другие.

Опубликована: Янв. 1, 2024

Язык: Английский

Global deep learning model for delineation of optically shallow and optically deep water in Sentinel-2 imagery DOI Creative Commons
Galen Richardson,

Neve Foreman,

Anders Knudby

и другие.

Remote Sensing of Environment, Год журнала: 2024, Номер 311, С. 114302 - 114302

Опубликована: Июль 4, 2024

In aquatic remote sensing, algorithms commonly used to map environmental variables rely on assumptions regarding the optical environment. Specifically, some assume that water is optically deep, i.e., influence of bottom reflectance measured signal negligible. Other opposite and are based an estimation bottom-reflected part signal. These may suffer from reduced performance when relevant not met. To address this, we introduce a general-purpose tool automates delineation deep shallow waters in Sentinel-2 imagery. This allows application for satellite-derived bathymetry, habitat identification, water-quality mapping be limited environments which they intended, thus enhance accuracy derived products. We sampled 440 images wide range coastal locations, covering all continents latitudes, manually annotated 1000 points each image as either or by visual interpretation. dataset was train six machine learning classification models - Maximum Likelihood, Random Forest, ExtraTrees, AdaBoost, XGBoost, neural networks utilizing both original top-of-atmosphere atmospherically corrected datasets. The were trained features including kernel means standard deviations band, well geographical location. A network emerged best model, with average 82.3% across two datasets fast processing time. Higher accuracies can achieved removing pixels intermediate probability scores predictions. made this model publicly available Python package. represents substantial step toward automatic imagery, sensing community downstream users ensure algorithms, such those bathymetry quality, applied only intended.

Язык: Английский

Процитировано

11
A global Swin-Unet Sentinel-2 surface reflectance-based cloud and cloud shadow detection algorithm for the NASA Harmonized Landsat Sentinel-2 (HLS) dataset DOI Creative Commons
Haiyan Huang, David P. Roy, Hugo De Lemos

и другие.

Science of Remote Sensing, Год журнала: 2025, Номер unknown, С. 100213 - 100213

Опубликована: Фев. 1, 2025

Язык: Английский

Процитировано

2
Global 30-m seamless data cube (2000–2022) of land surface reflectance generated from Landsat-5,7,8,9 and MODIS Terra constellations DOI Creative Commons
Shuang Chen, Jie Wang, Qiang Liu

и другие.

Опубликована: Июнь 17, 2024

Abstract. The Landsat series constitutes an unparalleled repository of multi-decadal Earth observations, serving as a cornerstone in global environmental monitoring. However, the inconsistent coverage data due to its long revisit intervals and frequent cloud cover poses significant challenges land monitoring over large geographical extents. In this study, we developed full-chain processing framework for multi-sensor fusion Landsat-5, 7, 8, 9 MODIS Terra surface reflectance products. Based on framework, global, 30-m resolution, daily Seamless Data Cube (SDC) was generated, spanning from 2000 2022. A thorough evaluation SDC undertaken using leave-one-out approach cross-comparison with NASA’s Harmonized Sentinel-2 (HLS) validation at 425 test sites assessed agreement between actual values (not used input), revealing overall Mean Absolute Error (MAE) 0.014 (the valid range is 0–1). HLS products 22 Military Grid Reference System (MGRS) tiles revealed Deviation (MAD) 0.017 L30 (Landsat-8-based product) MAD 0.021 S30 (Sentinel-2-based product). Moreover, experimental results underscore advantages employing classification, achieving sizable improvement accuracy (2.4 %~11.3 %) that obtained composite interpolated datasets. web-based interface has been researchers freely access dataset, which available https://doi.org/10.12436/SDC30.26.20240506 (Chen et al., 2024).

Язык: Английский

Процитировано

6
A computational framework for processing time-series of Earth Observation data based on discrete convolution: global-scale historical Landsat cloud-free aggregates at 30 m spatial resolution DOI Creative Commons
Davide Consoli, Leandro Parente, Rolf Simões

и другие.

Research Square (Research Square), Год журнала: 2024, Номер unknown

Опубликована: Май 24, 2024

Abstract Processing extremely large collections of Earth Observation (EO) time-series, often petabyte-sized, such as NASA's Landsat and ESA's Sentinel missions, can be computationally prohibitive costly. Despite their name, even the Analysis Ready Data (ARD) versions rarely used direct input for modeling require additional time-series processing. Existing solutions readily using these data are not openly available, poor in performance, or lack flexibility. Addressing this issue, we developed SIRCLE (Signal Imputation Refinement with Convolution Leaded Engine), a computational framework that to apply diverse processing techniques by simply adjusting convolution kernel. Together SIRCLE, paper presents SWAG (Seasonally Weighted Average Generalization), method EO reconstruction integrated framework. an imputation reconstruct images affected presence clouds. Compared benchmark dataset, consistently outperformed reference methods, reducing errors at least 15%. As first large-scale application, were employed process entire Global Land Discovery (GLAD) ARD-2 archive, producing cloud-free bi-monthly aggregated product. This process, covering seven bands globally from 1997 2022, more than two trillion pixels each one 156 samples product, required approximately 28 hours computation 1248 Intel(R) Xeon(R) Gold 6248R CPUs. The resulting reconstructed machine learning models map biophysical indices. With hosting about 20 TB per band/index 30 m resolution, historical stored Cloud-Optimized GeoTIFFs (COG) distributed open data, product enables seamless, fast, affordable access archive environmental monitoring analysis applications.

Язык: Английский

Процитировано

3
A computational framework for processing time-series of earth observation data based on discrete convolution: global-scale historical Landsat cloud-free aggregates at 30 m spatial resolution DOI Creative Commons
Davide Consoli, Leandro Parente, Rolf Simões

и другие.

PeerJ, Год журнала: 2024, Номер 12, С. e18585 - e18585

Опубликована: Дек. 4, 2024

Processing large collections of earth observation (EO) time-series, often petabyte-sized, such as NASA's Landsat and ESA's Sentinel missions, can be computationally prohibitive costly. Despite their name, even the Analysis Ready Data (ARD) versions rarely used direct input for modeling because cloud presence and/or storage size. Existing solutions readily using these data are not openly available, poor in performance, or lack flexibility. Addressing this issue, we developed TSIRF (Time-Series Iteration-free Reconstruction Framework), a computational framework that to apply diverse time-series processing tasks, temporal aggregation reconstruction by simply adjusting convolution kernel. As first large-scale application, was employed process entire Global Land Discovery (GLAD) ARD archive, producing cloud-free bi-monthly aggregated product. This process, covering seven bands globally from 1997 2022, with more than two trillion pixels each one 156 samples product, required approximately 28 hours computation 1248 Intel

Язык: Английский

Процитировано

3
Preface: Advancing deep learning for remote sensing time series data analysis DOI
Hankui K. Zhang, Gustau Camps‐Valls, Shunlin Liang

и другие.

Remote Sensing of Environment, Год журнала: 2025, Номер unknown, С. 114711 - 114711

Опубликована: Март 1, 2025

Язык: Английский

Процитировано

0
The Harmonized Landsat and Sentinel-2 version 2.0 surface reflectance dataset DOI Creative Commons

Junchang Ju,

Qiang Zhou, Brian Freitag

и другие.

Remote Sensing of Environment, Год журнала: 2025, Номер 324, С. 114723 - 114723

Опубликована: Апрель 16, 2025

Язык: Английский

Процитировано

0
Global 30 m seamless data cube (2000–2022) of land surface reflectance generated from Landsat 5, 7, 8, and 9 and MODIS Terra constellations DOI Creative Commons
Shuang Chen, Jie Wang, Qiang Liu

и другие.

Earth system science data, Год журнала: 2024, Номер 16(11), С. 5449 - 5475

Опубликована: Ноя. 28, 2024

Abstract. The Landsat series constitutes an unparalleled repository of multi-decadal Earth observations, serving as a cornerstone in global environmental monitoring. However, the inconsistent coverage data due to its long revisit intervals and frequent cloud cover poses significant challenges land monitoring over large geographical extents. In this study, we developed full-chain processing framework for multi-sensor fusion 5, 7, 8, 9 MODIS Terra surface reflectance products. Based on 30 m resolution daily seamless cube (SDC) was generated, spanning from 2000 2022. A thorough evaluation SDC undertaken using leave-one-out approach cross-comparison with NASA's Harmonized Sentinel-2 (HLS) validation at 425 test sites assessed agreement between actual values (not used input), revealing overall mean absolute error (MAE) 0.014 (the valid range is 0–1). HLS products 22 Military Grid Reference System (MGRS) tiles revealed deviation (MAD) 0.017 L30 (Landsat 8-based product) MAD 0.021 S30 (Sentinel-2-based product). Moreover, experimental results underscore advantages employing classification, achieving sizable improvement accuracy (2.4 %–11.3 %) that obtained composite interpolated datasets. web-based interface has been researchers freely access dataset, which available https://doi.org/10.12436/SDC30.26.20240506 (Chen et al., 2024).

Язык: Английский

Процитировано

3
Quantitative spatiotemporal evolution of large urban agglomeration expansion based on 1995–2020 nighttime light and spectral data DOI Creative Commons
Yuanmao Zheng,

Yaling Cai,

Kexin Yang

и другие.

Ecological Informatics, Год журнала: 2024, Номер unknown, С. 102824 - 102824

Опубликована: Сен. 1, 2024

Язык: Английский

Процитировано

2
XDGGS: A community-developed Xarray package to support planetary DGGS data cube computations DOI Creative Commons
Alexander Kmoch,

Benoît Bovy,

Justus Magin

и другие.

˜The œinternational archives of the photogrammetry, remote sensing and spatial information sciences/International archives of the photogrammetry, remote sensing and spatial information sciences, Год журнала: 2024, Номер XLVIII-4/W12-2024, С. 75 - 80

Опубликована: Июнь 20, 2024

Abstract. Traditional map projections introduce distortions, especially for global data. Discrete Global Grid Systems (DGGS) offer an alternative by dividing the Earth into equal-area grid cells at different resolutions. This paper describes xdggs, a new Xarray extension that simplifies working with DGGS. Xdggs provides unified API various DGGS libraries and integrates seamlessly Pangeo ecosystem through extending widely used library to use DGGS-specific cell identifiers as index. development makes more accessible will lead facilitating data analysis on planetary scale.Xdggs aims provide user-friendly hides implementation complexities of libraries. And because it Xarray, popular tool geospatial analysis, xdggs promotes FAIR practices simplifying access interoperability can become valuable scientists application developers datasets.

Язык: Английский

Процитировано

1