scholarly journals A COMPARISON OF TREE-BASED ALGORITHMS FOR COMPLEX WETLAND CLASSIFICATION USING THE GOOGLE EARTH ENGINE

2021 ◽  
Vol XLVI-4/W5-2021 ◽  
pp. 313-319
Author(s):  
A. Jamali ◽  
M. Mahdianpari ◽  
İ. R. Karaş
Keyword(s):  
Remote Sensing ◽  
Large Scale ◽  
Well Being ◽  
Google Earth ◽  
The Other ◽  
Gradient Boosting ◽  
Wetland Classification ◽  
Observation Methods ◽  
Extreme Gradient Boosting ◽  
Google Earth Engine

Abstract. Wetlands are endangered ecosystems that are required to be systematically monitored. Wetlands have significant contributions to the well-being of human-being, fauna, and fungi. They provide vital services, including water storage, carbon sequestration, food security, and protecting the shorelines from floods. Remote sensing is preferred over the other conventional earth observation methods such as field surveying. It provides the necessary tools for the systematic and standardized method of large-scale wetland mapping. On the other hand, new cloud computing technologies for the storage and processing of large-scale remote sensing big data such as the Google Earth Engine (GEE) have emerged. As such, for the complex wetland classification in the pilot site of the Avalon, Newfoundland, Canada, we compare the results of three tree-based classifiers of the Decision Tree (DT), Random Forest (RF), and Extreme Gradient Boosting (XGB) available in the GEE code editor using Sentinel-2 images. Based on the results, the XGB classifier with an overall accuracy of 82.58% outperformed the RF (82.52%) and DT (77.62%) classifiers.

scholarly journals High-Resolution Vegetation Mapping Using eXtreme Gradient Boosting Based on Extensive Features

2019 ◽  
Vol 11 (12) ◽  
pp. 1505 ◽  
Author(s):  
Heng Zhang ◽  
Anwar Eziz ◽  
Jian Xiao ◽  
Shengli Tao ◽  
Shaopeng Wang ◽  
...  
Keyword(s):  
New Zealand ◽  
High Resolution ◽  
Large Scale ◽  
Field Survey ◽  
Terrestrial Ecosystems ◽  
Google Earth ◽  
Gradient Boosting ◽  
Global Changes ◽  
Vegetation Map ◽  
Extreme Gradient Boosting

Accurate mapping of vegetation is a premise for conserving, managing, and sustainably using vegetation resources, especially in conditions of intensive human activities and accelerating global changes. However, it is still challenging to produce high-resolution multiclass vegetation map in high accuracy, due to the incapacity of traditional mapping techniques in distinguishing mosaic vegetation classes with subtle differences and the paucity of fieldwork data. This study created a workflow by adopting a promising classifier, extreme gradient boosting (XGBoost), to produce accurate vegetation maps of two strikingly different cases (the Dzungarian Basin in China and New Zealand) based on extensive features and abundant vegetation data. For the Dzungarian Basin, a vegetation map with seven vegetation types, 17 subtypes, and 43 associations was produced with an overall accuracy of 0.907, 0.801, and 0.748, respectively. For New Zealand, a map of 10 habitats and a map of 41 vegetation classes were produced with 0.946, and 0.703 overall accuracy, respectively. The workflow incorporating simplified field survey procedures outperformed conventional field survey and remote sensing based methods in terms of accuracy and efficiency. In addition, it opens a possibility of building large-scale, high-resolution, and timely vegetation monitoring platforms for most terrestrial ecosystems worldwide with the aid of Google Earth Engine and citizen science programs.

scholarly journals MAPPING OF BURNED AREA USING PRESENCE AND BACKGROUND LEARNING FRAMEWORK ON THE GOOGLE EARTH ENGINE PLATFORM

2019 ◽  
Vol XLII-3/W8 ◽  
pp. 37-41
Author(s):  
D. Attaf ◽  
K. Djerriri ◽  
D. Mansour ◽  
D. Hamdadou
Keyword(s):  
Remote Sensing ◽  
Classification Problem ◽  
Google Earth ◽  
The Other ◽  
Burned Area ◽  
Support Vector ◽  
Classification Techniques ◽  
Burned Areas ◽  
Google Earth Engine ◽  
One Class Classification

<p><strong>Abstract.</strong> Mapping of burned areas caused by forest fires was always a main concern to researchers in the field of remote sensing. Thus, various spectral indices and classification techniques have been proposed in the literature. In such a problem, only one specific class is of real interest and could be referred to as a one-class classification problem. One-class classification methods are highly desirable for quick mapping of classes of interest. A common used solution to deal with One-Class classification problem is based on oneclass support vector machine (OC-SVM). This method has proved useful in classification of remote sensing images. However, overfitting problem and difficulty in tuning parameters have become the major obstacles for this method. The new Presence and Background Learning (PBL) framework does not require complicated model selection and can generate very high accuracy results. On the other hand the Google Earth Engine (GEE) portal provides access to satellite and other ancillary data, cloud computing, and algorithms for processing large amounts of data with relative ease. Therefore, this study mainly aims to investigate the possibility of using the PBL framework within the GEE platform to extract burned areas from freely available Landsat archive in the year 2015. The quality of the results obtained using PBL framework was assessed using ground truth digitized by qualified technicians and compared to other classification techniques: Thresholding burned area spectral Index (BAI) and OC-SVM classifiers. Experimental results demonstrate that PBL framework for mapping the burned areas shows the higher classification accuracy than the other classifiers, and it highlights the suitability for the cases with few positive labelled samples available, which facilitates the tedious work of manual digitizing.</p>

scholarly journals The NDVI algorithm utilization on the google earth engine platform to monitor changes in forest density in mining area

2021 ◽  
Vol 886 (1) ◽  
pp. 012100
Author(s):  
Munajat Nursaputra ◽  
Siti Halimah Larekeng ◽  
Nasri ◽  
Andi Siady Hamzah
Keyword(s):  
Remote Sensing ◽  
Time Series ◽  
Large Scale ◽  
Vegetation Index ◽  
Forest Degradation ◽  
Google Earth ◽  
Aerial Images ◽  
Forest Monitoring ◽  
Forest Density ◽  
Google Earth Engine

Abstract Periodic forest monitoring needs to be done to avoid forest degradation. In general, forest monitoring can be conducted manually (field surveys) or using technological innovations such as remote sensing data derived from aerial images (drone results) or cloud computing-based image processing. Currently, remote sensing technology provides large-scale forest monitoring using multispectral sensors and various vegetation index processing algorithms. This study aimed to evaluate the use of the Google Earth Engine (GEE) platform, a geospatial dataset platform, in the Vale Indonesia mining concession area to improve accountable forest monitoring. This platform integrates a set of programming methods with a publicly accessible time-series database of satellite imaging services. The method used is NDVI processing on Landsat multispectral images in time series format, which allows for the description of changes in forest density levels over time. The results of this NDVI study conducted on the GEE platform have the potential to be used as a tool and additional supporting data for monitoring forest conditions and improvement in mining regions.

scholarly journals Rapid Mapping of Large-Scale Greenhouse Based on Integrated Learning Algorithm and Google Earth Engine

2021 ◽  
Vol 13 (7) ◽  
pp. 1245
Author(s):  
Jinhuang Lin ◽  
Xiaobin Jin ◽  
Jie Ren ◽  
Jingping Liu ◽  
Xinyuan Liang ◽  
...  
Keyword(s):  
Remote Sensing ◽  
Agricultural Production ◽  
Large Scale ◽  
Texture Features ◽  
Google Earth ◽  
Distribution Data ◽  
Classification Algorithms ◽  
Google Earth Engine ◽  
Remote Sensing Mapping ◽  
Terrain Features

A greenhouse is an important land-use type, which can effectively improve agricultural production conditions and increase crop yields. It is of great significance to obtain the spatial distribution data of greenhouses quickly and accurately for regional agricultural production and food security. Based on the Google Earth Engine cloud platform and Landsat 8 images, this study selected a total of 18 indicators from three aspects of spectral features, texture features and terrain features to construct greenhouse identification features. From a variety of classification algorithms for remote-sensing recognition of greenhouses, this study selected three classifiers with higher accuracy (classification and regression trees (CART), random forest model (randomForest) and maximum entropy model (gmoMaxEnt)) to construct an integrated classification algorithm, and then extracted the spatial distribution data of greenhouses in Jiangsu Province. The results show that: (1) Google Earth Engine with its own massive data and cloud computing capabilities, combined with integrated classification algorithms, can achieve rapid remote-sensing mapping of large-scale greenhouses under complex terrain, and the classification accuracy is higher than that of a single classification algorithm. (2) The combination of different spectral, texture and terrain features has a greater impact on the extraction of regional greenhouses, the combination of all three aspects of features has the highest accuracy. Spectral features are the key factors for greenhouse remote-sensing mapping, but terrain and texture features can also enhance classification accuracy. (3) The greenhouse in Jiangsu Province has significant spatial differentiation and spatial agglomeration characteristics. The most widely distributed greenhouses are mainly concentrated in the agriculturally developed areas such as Dongtai City, Hai’an County, Rudong County and Pizhou City.

scholarly journals The Google Earth Engine Digitisation Tool (GEEDiT) and the Margin change Quantification Tool (MaQiT) – simple tools for the rapid mapping and quantification of changing Earth surface margins

2018 ◽  
Vol 6 (3) ◽  
pp. 551-561 ◽  
Author(s):  
James M. Lea
Keyword(s):  
Remote Sensing ◽  
High School Students ◽  
Large Scale ◽  
Research Question ◽  
Google Earth ◽  
Earth Surface ◽  
School Students ◽  
Glacier Terminus ◽  
Wide Range ◽  
Google Earth Engine

Abstract. Changes in margins derived from satellite imagery are quantitative indicators of the environmental processes and drivers acting on the Earth's surface, for example retreating ice margins or coastal changes with rising sea level. However, the large-scale rapid visualisation and analysis of the satellite record is often impractical due to factors such as computer processing power, software availability, internet connection speed and/or user expertise in remote sensing. Here are presented three new, freely accessible tools that together can be used to process, visualise and review data from the full Landsat 4–8 and Sentinel 1–2 satellite records in seconds, enabling efficient mapping (through manual digitisation) and automated quantification of margin changes. These tools are highly accessible for users from a range of remote-sensing expertise (from academics to high school students), with minimal computational, licensing and knowledge-based barriers to access. The Google Earth Engine Digitisation Tool (GEEDiT) allows users to define a point anywhere on the planet and filter data from each satellite for user-defined time frames, maximum acceptable cloud cover extent, and options of predefined or custom image band combinations via a simple graphical user interface (GUI). GEEDiT allows georeferenced vectors to be easily and rapidly mapped from each image, with image metadata and user notes automatically appended to each vector, which can then be exported for subsequent analysis. The GEEDiT Reviewer tool allows users to quality control their own/others' data and also filter existing datasets based on the spatial/temporal requirements for their particular research question. The Margin change Quantification Tool (MaQiT) is complementary to GEEDiT and GEEDiT Reviewer, allowing the rapid quantification of these margin changes by utilising two well-established methods that have previously been used to measure glacier margin change and two new methods via a similarly simple GUI. A case study of the lake-terminating glacier Breiðamerkurjökull, Iceland, is used to demonstrate the complementary functionality of GEEDiT, GEEDiT Reviewer and MaQiT, though it should be noted that MaQiT is also suitable for the (re-)analysis of existing datasets not generated by GEEDiT. MaQiT has been developed with the original aim of quantifying tidewater glacier terminus change, though the methods included within the tool have potential for wide applications in multiple areas of Earth surface science (e.g. coastal and vegetation extent change). It is hoped that these tools will allow a wide range of researchers and students across the geosciences to efficiently map, analyse and access volumes of data that would have previously proven prohibitive.

Continent-wide bimonthly mapping of Antarctic surface meltwater using Google Earth Engine

2021 ◽  
Author(s):  
Peter Tuckett ◽  
Jeremy Ely ◽  
Andrew Sole ◽  
Stephen Livingstone ◽  
James Lea
Keyword(s):  
Remote Sensing ◽  
Spatial Analysis ◽  
Mass Balance ◽  
Large Scale ◽  
Ice Sheet ◽  
Google Earth ◽  
Landsat 8 ◽  
Supraglacial Lakes ◽  
Google Earth Engine ◽  
Temporal And Spatial

&lt;p&gt;Surface meltwater is widespread around the margin of the Antarctic Ice Sheet during the austral summer. This meltwater, typically transported via surface streams and rivers and stored in supraglacial lakes, has the potential to influence ice-sheet mass balance through ice-dynamic and albedo feedbacks. To predict the impact that surface melt will have on mass balance over coming decades, it is important to understand spatial and temporal variability in surface meltwater extent. A variety of methods have been used to detect supraglacial lakes in Antarctica, yet a multi-annual, continent-wide study of Antarctic supraglacial meltwater has yet to be conducted. Cloud-based computational platforms, such as Google Earth Engine (GEE), enable large-scale temporal and spatial analysis of remote sensing datasets at minimal time expense. Here, we implement an automated method for meltwater detection in GEE to generate continent-wide, bimonthly repeat assessments of supraglacial lake extent between 2013 and 2020. We use a band-threshold based approach to delineate surface water from Landsat-8 imagery. Furthermore, our method incorporates a novel technique for quantifying meltwater extent that accounts for variability in optical image coverage and cloud cover, enabling an upper uncertainty bound to be attached to minimum mapped lake areas. We present results from continent-wide mapping, and highlight initial findings that indicate evolution of lakes in Antarctica over the past seven years. This work demonstrates how platforms such as GEE have revolutionized our ability to undertake large-scale projects from remote sensing datasets, allowing for greater temporal and spatial analysis of cryospheric processes than previously possible.&lt;/p&gt;

scholarly journals Remote Sensing Based Quantification of Forest Cover Change in Somalia for the Period 2000 to 2019

2021 ◽  
Author(s):  
Sylus Kipngeno Musei ◽  
Justine Muhoro Nyaga ◽  
Abdi Zeila Dubow
Keyword(s):  
Remote Sensing ◽  
Land Degradation ◽  
Large Scale ◽  
Forest Cover ◽  
Computing System ◽  
Google Earth ◽  
Forest Trees ◽  
Forest Cover Change ◽  
Charcoal Production ◽  
Google Earth Engine

Deforestation is a driver of land degradation and a major environmental problem in Somalia, and has been linked to frequent incidences of drought over the years. Monitoring of changes in forest cover is therefore critical for the country’s environment. The problem of land degradation has been worsened by the large scale charcoal production that is witnessed in the country. This study aimed at estimating forest cover change between 2000 and 2019 in Somalia using Landsat-based forest cover datasets. Google Earth Engine (GEE), a cloud based computing system was used to provide a platform for this analysis. Based on the 30% threshold recommended by International Geosphere Biosphere Program for differentiating forest from non-forest trees, approximately 23% forest cover loss was found, from 87, 294 hectares in 2000 to 67, 199 hectares in 2019. Most of the country’s forest is within the southern and central parts of the country, and significant forest cover losses occurred mainly around Mogadishu and Kismayo port throughout the study period. There is therefore a need for the Federal Ministry of Environment and environment ministries in the federal member states to design mechanisms and strategies for restoration of the degraded forests and to curb deforestation.

scholarly journals Google Earth Engine Digitisation Tool (GEEDiT), and Margin change Quantification Tool (MaQiT) – simple tools for the rapid mapping and quantification of changing Earth surface margins

2018 ◽  
Author(s):  
James M. Lea
Keyword(s):  
Remote Sensing ◽  
Surface Science ◽  
Large Scale ◽  
Google Earth ◽  
Earth Surface ◽  
Boundary Change ◽  
Glacier Terminus ◽  
Processing Power ◽  
Wide Range ◽  
Google Earth Engine

Abstract. The visualisation and exploration of satellite imagery archives coupled with the quantification of margin/boundary changes are frequently used within earth surface sciences as key indicators of the environmental processes and drivers acting within a system. However, the large scale rapid visualisation and analysis of this imagery is often impractical due to factors such as computer processing power, software availability, internet connection speed, and user expertise in remote sensing. Here are described two separate tools that together can be used to process and visualise the full Landsat 4–8 and Sentinel 1–2 satellite records in seconds, enabling efficient mapping (through manual digitisation) and automated quantification of margin changes. These tools are highly accessible for users from a range of remote sensing expertise, with minimal computational, licensing and knowledge-based barriers to access. The Google Earth Engine Digitisation Tool (GEEDiT) allows users to define a point anywhere on the planet and access all Landsat 4–8/Sentinel 1–2 imagery at that location, filtered for user defined time frames, maximum acceptable cloud cover extent, and options of predefined or custom image band combinations via a simple Graphical User Interface (GUI). GEEDiT also allows georeferenced vectors to be easily and rapidly mapped from each image with image metadata and user notes automatically appended to each vector. This data can then be exported to a user's Google Drive for subsequent analysis. The Margin change Quantification Tool (MaQiT) is complimentary to GEEDiT, allowing the rapid quantification of these margin changes utilising two well-established methods that have previously been used to measure glacier margin change and two new methods via a similarly simple GUI. MaQiT is also suitable for the (re-)analysis of existing datasets not generated by GEEDiT. Although MaQiT has been developed with the aim of quantifying tidewater glacier terminus change, the tool can be applied to other margin changes within earth surface science where margin/boundary change through time is of interest (e.g. coastal and vegetation extent change). It is hoped that these tools will allow a wide range of researchers and students across the geosciences to have access to, efficiently map and analyse volumes of data that may have previously proven prohibitive.

scholarly journals Multi-Temporal Land Cover Change Mapping Using Google Earth Engine and Ensemble Learning Methods

2020 ◽  
Vol 10 (22) ◽  
pp. 8083 ◽  
Author(s):  
Nimisha Wagle ◽  
Tri Dev Acharya ◽  
Venkatesh Kolluru ◽  
He Huang ◽  
Dong Ha Lee
Keyword(s):  
Land Cover ◽  
Ensemble Learning ◽  
Vegetation Index ◽  
Confusion Matrix ◽  
Google Earth ◽  
Gradient Boosting ◽  
Observation Data ◽  
Learning Methods ◽  
Extreme Gradient Boosting ◽  
Google Earth Engine

The study deals with the application of Google Earth Engine (GEE), Landsat data and ensemble-learning methods (ELMs) to map land cover (LC) change over a decade in the Kaski district of Nepal. As Nepal has experienced extensive changes due to natural and anthropogenic activities, monitoring such changes are crucial for understanding relationships and interactions between social and natural phenomena and to promote better decision-making. The main novelty lies in applying the XGBoost classifier for LC mapping over Nepal and monitoring the decadal changes of LC using ELMs. To map the LC change, a yearly cloud-free composite Landsat image was selected for the year 2010 and 2020. Combining the annual normalized difference vegetation index, normalized difference built-up index and modified normalized difference water index, with elevation and slope data from shuttle radar topography mission, supervised classification was performed using a random forest and extreme gradient boosting ELMs. Post classification change detection, validation and accuracy assessment were executed after the preparation of the LC maps. Three evaluation indices, namely overall accuracy (OA), Kappa coefficient, and F1 score from confusion matrix reports, were calculated for all the points used for validation purposes. We have obtained an OA of 0.8792 and 0.875 for RF and 0.8926 and 0.8603 for XGBoost at the 95% confidence level for 2010 and 2020 LC maps, which are better for mountainous terrain. The applied methodology could be significant in utilizing the big earth observation data and overcoming the traditional computational challenges using GEE. In addition, the quantification of changes over time would be helpful for decision-makers to understand current environmental dynamics in the study area.

Sign in / Sign up

Export Citation Format

Share Document