scholarly journals Assessment of Annual Composite Images Obtained by Google Earth Engine for Urban Areas Mapping Using Random Forest

2021 ◽  
Vol 13 (4) ◽  
pp. 748
Author(s):  
Zhaoming Zhang ◽  
Mingyue Wei ◽  
Dongchuan Pu ◽  
Guojin He ◽  
Guizhou Wang ◽  
...  
Keyword(s):  
Time Series ◽  
Random Forest ◽  
Urban Areas ◽  
Google Earth ◽  
Landsat 8 ◽  
Annual Time Series ◽  
Google Earth Engine ◽  
Time Series Images ◽  
The Impact ◽  
Annual Time

Urban areas represent the primary source region of greenhouse gas emissions. Mapping urban areas is essential for understanding land cover change, carbon cycles, and climate change (urban areas also refer to impervious surfaces, i.e., artificial cover and structures). Remote sensing has greatly advanced urban areas mapping over the last several decades. At present, we have entered the era of big data. Long time series of satellite data such as Landsat and high-performance computing platforms such as Google Earth Engine (GEE) offer new opportunities to map urban areas. The objective of this research was to determine how annual time series images from Landsat 8 Operational Land Imager (OLI) can effectively be composed to map urban areas in three cities in China in support of GEE. Three reducer functions, ee.Reducer.min(), ee.Reducer.median(), and ee.Reducer.max() provided by GEE, were selected to construct four schemes to synthesize the annual intensive time series Landsat 8 OLI data for three cities in China. Then, urban areas were mapped based on the random forest algorithm and the accuracy was evaluated in detail. The results show that (1) the quality of annual composite images was improved significantly, particularly in reducing the impact of cloud and cloud shadows, and (2) the annual composite images obtained by the combination of multiple reducer functions had better performance than that obtained by a single reducer function. Further, the overall accuracy of urban areas mapping with the combination of multiple reducer functions exceeded 90% in all three cities in China. In summary, a suitable combination of reducer functions for synthesizing annual time series images can enhance data quality and ensure differences between characteristics and higher precision for urban areas mapping.

scholarly journals Land Cover Classification using Google Earth Engine and Random Forest Classifier—The Role of Image Composition

2020 ◽  
Vol 12 (15) ◽  
pp. 2411 ◽  
Author(s):  
Thanh Noi Phan ◽  
Verena Kuch ◽  
Lukas W. Lehnert
Keyword(s):  
Time Series ◽  
Random Forest ◽  
Land Cover ◽  
Land Cover Classification ◽  
High Accuracy ◽  
Google Earth ◽  
Temporal Aggregation ◽  
Google Earth Engine ◽  
Time Series Images ◽  
Land Cover Maps

Land cover information plays a vital role in many aspects of life, from scientific and economic to political. Accurate information about land cover affects the accuracy of all subsequent applications, therefore accurate and timely land cover information is in high demand. In land cover classification studies over the past decade, higher accuracies were produced when using time series satellite images than when using single date images. Recently, the availability of the Google Earth Engine (GEE), a cloud-based computing platform, has gained the attention of remote sensing based applications where temporal aggregation methods derived from time series images are widely applied (i.e., the use the metrics such as mean or median), instead of time series images. In GEE, many studies simply select as many images as possible to fill gaps without concerning how different year/season images might affect the classification accuracy. This study aims to analyze the effect of different composition methods, as well as different input images, on the classification results. We use Landsat 8 surface reflectance (L8sr) data with eight different combination strategies to produce and evaluate land cover maps for a study area in Mongolia. We implemented the experiment on the GEE platform with a widely applied algorithm, the Random Forest (RF) classifier. Our results show that all the eight datasets produced moderately to highly accurate land cover maps, with overall accuracy over 84.31%. Among the eight datasets, two time series datasets of summer scenes (images from 1 June to 30 September) produced the highest accuracy (89.80% and 89.70%), followed by the median composite of the same input images (88.74%). The difference between these three classifications was not significant based on the McNemar test (p > 0.05). However, significant difference (p < 0.05) was observed for all other pairs involving one of these three datasets. The results indicate that temporal aggregation (e.g., median) is a promising method, which not only significantly reduces data volume (resulting in an easier and faster analysis) but also produces an equally high accuracy as time series data. The spatial consistency among the classification results was relatively low compared to the general high accuracy, showing that the selection of the dataset used in any classification on GEE is an important and crucial step, because the input images for the composition play an essential role in land cover classification, particularly with snowy, cloudy and expansive areas like Mongolia.

scholarly journals The impact of forest regeneration on streamflow in 12 mesoscale humid tropical catchments

2013 ◽  
Vol 17 (7) ◽  
pp. 2613-2635 ◽  
Author(s):  
H. E. Beck ◽  
L. A. Bruijnzeel ◽  
A. I. J. M. van Dijk ◽  
T. R. McVicar ◽  
F. N. Scatena ◽  
...  
Keyword(s):  
Time Series ◽  
Land Cover ◽  
Forest Regeneration ◽  
Forest Cover ◽  
Impact Analysis ◽  
Forest Area ◽  
Macro Scale ◽  
Annual Time Series ◽  
The Impact ◽  
Annual Time

Abstract. Although regenerating forests make up an increasingly large portion of humid tropical landscapes, little is known of their water use and effects on streamflow (Q). Since the 1950s the island of Puerto Rico has experienced widespread abandonment of pastures and agricultural lands, followed by forest regeneration. This paper examines the possible impacts of these secondary forests on several Q characteristics for 12 mesoscale catchments (23–346 km2; mean precipitation 1720–3422 mm yr−1) with long (33–51 yr) and simultaneous records for Q, precipitation (P), potential evaporation (PET), and land cover. A simple spatially-lumped, conceptual rainfall–runoff model that uses daily P and PET time series as inputs (HBV-light) was used to simulate Q for each catchment. Annual time series of observed and simulated values of four Q characteristics were calculated. A least-squares trend was fitted through annual time series of the residual difference between observed and simulated time series of each Q characteristic. From this the total cumulative change (Â) was calculated, representing the change in each Q characteristic after controlling for climate variability and water storage carry-over effects between years. Negative values of  were found for most catchments and Q characteristics, suggesting enhanced actual evaporation overall following forest regeneration. However, correlations between changes in urban or forest area and values of  were insignificant (p &amp;geq; 0.389) for all Q characteristics. This suggests there is no convincing evidence that changes in the chosen Q characteristics in these Puerto Rican catchments can be ascribed to changes in urban or forest area. The present results are in line with previous studies of meso- and macro-scale (sub-)tropical catchments, which generally found no significant change in Q that can be attributed to changes in forest cover. Possible explanations for the lack of a clear signal may include errors in the land cover, climate, Q, and/or catchment boundary data; changes in forest area occurring mainly in the less rainy lowlands; and heterogeneity in catchment response. Different results were obtained for different catchments, and using a smaller subset of catchments could have led to very different conclusions. This highlights the importance of including multiple catchments in land-cover impact analysis at the mesoscale.

scholarly journals Mapeamento da Vegetação Nativa do Cerrado na Região de Três Lagoas-MS com o Google Earth Engine

2019 ◽  
Vol 71 (3) ◽  
pp. 702-725
Author(s):  
Nayara Vasconcelos Estrabis ◽  
José Marcato Junior ◽  
Hemerson Pistori
Keyword(s):  
Support Vector Machine ◽  
Random Forest ◽  
Google Earth ◽  
Support Vector ◽  
Landsat 8 ◽  
Landsat 8 Oli ◽  
Google Earth Engine

O Cerrado é um dos biomas existentes no Brasil e o segundo mais extenso da América do Sul. Possui grande importância devido a sua biodiversidade, ecossistema e principalmente por servir como um reservatório, ou “esponja”, que distribui água para os demais biomas, além de ser berço de nascentes de algumas das maiores bacias da América do Sul. No entanto, devido às atividades antrópicas praticadas (com destaque para a pecuária e silvicultura) e a redução da vegetação nativa, este bioma está ameaçado. Considerado como hotspot em biodiversidade, o Cerrado pode não existir em 2050. Com a necessidade de sua preservação, o objetivo desse trabalho consistiu em investigar o uso de algoritmos de aprendizado de máquina para realizar o mapeamento da vegetação nativa existente na região do município de Três Lagoas, utilizando a plataforma em nuvem Google Earth Engine. O processo foi realizado com uma imagem Landsat-8 OLI, datada de 10 de outubro de 2018, e com os algoritmos Random Forest (RF) e Support Vector Machine (SVM). Na validação da classificação, o RF e o SVM apresentaram índices kappa iguais a 0,94 e 0,97, respectivamente. O RF, quando comparado ao SVM, apresentou classificação mais ruidosa. Por fim, verificou-se a existência de vegetação nativa de aproximadamente 2556 km² ao adotar o RF e 2873 km² ao adotar SVM.

scholarly journals DETECÇÃO DE ÁREAS DE FLORESTAS INVARIANTES EM SÉRIES TEMPORAIS UTILIZANDO RANDOM FOREST

GEOgraphia ◽  
2021 ◽  
Vol 23 (50) ◽  
Author(s):  
Eduardo Ribeiro Lacerda ◽  
Raúl Sanchéz Vicens
Keyword(s):  
Land Use ◽  
Time Series ◽  
Random Forest ◽  
Change Detection ◽  
Vegetation Change ◽  
Google Earth ◽  
Detection Algorithms ◽  
Palabras Clave ◽  
Change Detection Algorithms ◽  
Google Earth Engine

O surgimento de algoritmos de detecção de mudanças na vegetação na última década é impressionante. Mas os resultados gerados ainda possuem ruído que precisa ser tratado com a utilização de resultados de outros mapeamentos de cobertura vegetal. Além disso, a necessidade de gerar classes de uso do solo invariantes é importante para o melhor entendimento de processos que ocorrem em áreas florestais. Pensando nisso, este trabalho busca criar uma nova forma de mapear essas áreas invariáveis que possam ser utilizadas para mascarar ruídos e também como subsídio para outros estudos de conservação e restauração. A metodologia proposta aqui usa a plataforma Google Earth Engine e um algoritmo de aprendizado de máquina: o Random Forest, para classificar áreas de floresta invariáveis usando todo o acervo de imagens da série temporal Landsat, de uma só vez. Os resultados mostraram que a nova abordagem teve melhor desempenho do que o uso de técnicas mais tradicionais como a agregação de mapeamentos de uso e cobertura anuais, com uma acurácia global de 91,7%. O trabalho busca ainda contribuir com a comunidade de sensoriamento remoto ao apresentar, após exaustivos testes, as melhores opções de variáveis a serem utilizadas neste tipo de classificação. Palavras-chave: Séries Temporais, Detecção de Mudanças, Florestas, Google Earth Engine, Random Forest.DETECTION OF INVARIANT VEGETATION AREAS IN TIME SERIES USING RANDOM FOREST ALGORITHMAbstract: The emergence of vegetation change detection algorithms in the last decade is impressive. But the results still have a lot of noise that needs to be cleaned. And the data cleaning process still uses other landcover mapping results. Besides that, the necessity to generate invariant land use classes is important to know particularly to forest areas. Thinking about that, this paper seeks to create a new form of mapping these invariant areas that can be used to mask noise and as an input on other conservation and restoration studies. The methodology proposed here uses the Google Earth Engine platform and a Random Forest algorithm to classify invariant forest areas using all the image’s collection in the time series at once. The results showed that the new approach performed better than the use of more traditional techniques such as the aggregation of annual land-use and land-cover mappings, with an overall accuracy of 91.7%. Also, this paper seeks to contribute to the remote sensing community showing after exhaustive testing, good options of variables to use on this type of work. Keywords: Time Series, Change Detection, Forests, Google Earth Engine, Random Forest.DETECCIÓN DE ÁREAS DE VEGETACIÓN INVARIANTES EN SÉRIES TEMPORALES UTILIZANDO ALGORITMO RANDOM FORESTResumen: La aparición de algoritmos de detección de cambios en la vegetación en la última década es impresionante. Pero los resultados todavía tienen muchos ruidos que deben ser eliminados. Además, el proceso de limpieza de datos se basa en otros mapas de cobertura de la tierra. Además de eso, es importante conocer la necesidad de generar clases de uso de la tierra invariables, particularmente en las áreas forestales. Pensando en eso, este artículo busca crear una nueva forma de mapear estas áreas invariantes que se pueden utilizar para enmascarar el ruido y como un aporte para otros estudios de conservación y restauración. La metodología propuesta aquí utiliza la plataforma Google Earth Engine y un algoritmo de aprendizaje de máquina: o Random Forest para clasificar áreas invariantes de bosque, utilizando a la vez todas las imágenes de la serie temporal Landsat. Los resultados encontraron que el nuevo enfoque tuvo mejor desempeño que el uso de técnicas tradicionales, con una precisión global del 91,7%. Este trabajo busca además contribuir con la comunidad de la teledetección, mostrando mediante de exhaustivas pruebas, mejores opciones de variables para utilizar en este tipo de clasificación. Palabras clave: Series de Tiempo, Detección de Cambios, Bosques, Google Earth Engine, Random Forest.

scholarly journals SPATIOTEMPORAL CHANGE ANALYSIS OF THE PROTECTED AREAS: A CASE STUDY – İĞNEADA FLOODPLAIN FORESTS

2021 ◽  
Vol XLIII-B3-2021 ◽  
pp. 735-740
Author(s):  
M. Toker ◽  
E. Çolak ◽  
F. Sunar
Keyword(s):  
Time Series ◽  
Protected Areas ◽  
National Parks ◽  
National Park ◽  
Spatial Scales ◽  
Google Earth ◽  
Floodplain Forests ◽  
Landsat Satellite ◽  
Time Series Images ◽  
The Impact

Abstract. Protected areas are important with land or water body ecosystems that have biodiversity, flora and fauna species. In Turkey, National Parks are one of the protected areas managed according to the National Parks Law No. 2873. Among them, the İğneada Floodplain Forests National Park, located in İğneada town in the province of Kırklareli, Turkey has been declared as a national park in 2007, and has an importance being a rare ecosystem, which consists of wetland, swamp, lakes and coastal sand dunes. Planning of Protected Areas can be done in a variety of ways, taking into account the balance of protection/use and should follow policies and guidelines. Today, for the sustainability and effective management of forest ecosystems, remote sensing technology provides an effective tool for assessing and monitoring ecosystem health at different temporal and spatial scales. In this study, potential temporal changes in the National Park were analyzed with Landsat satellite time series images using two different methods. First method, the Landtrendr algorithm (Landsat-based Detection of Trends in Disturbance and Recovery) developed for multitemporal satellite data, uses pixel values as input data and analysis them by using regression models to capture, label and map the changes. In this context, Landsat satellite time series images were taken quinquennial between 1987 and 2007 and biennially until 2017 for Landtrendr analysis (i.e. before and after its declaration as a National Park, respectively). As a second approach, the Google Earth Engine (GEE) cloud-based platform, which facilitates access to high-performance computing resources to process large long-term data sets, was used to analyze the impact of land cover changes. The results showed that the area was subjected to various pressures (i.e. due to illegal felling, pollution, etc.) until it was declared as a National park. Although there was general improvement and recovery after the region declared as a Park, it was seen that the sensitive dynamics of the region require continuous monitoring and protection using geo-information technologies.

scholarly journals BUILDING OF TEMPERATURE MAPS OF NOVOSIBIRSK CITY BASED ON LANDSAT 8 DATA

2021 ◽  
Vol 4 (1) ◽  
pp. 52-59
Author(s):  
Elena A. Mamash ◽  
Igor A. Pestunov ◽  
Dmitrii L. Chubarov
Keyword(s):  
Time Series ◽  
Satellite Data ◽  
Google Earth ◽  
Underlying Surface ◽  
Landsat 8 ◽  
Google Earth Engine ◽  
Engine System ◽  
Temperature Maps

An algorithm for constructing temperature maps of the underlying surface based on a multi-time series of atmospheric corrected satellite data from Landsat 8, implemented in the Google Earth Engine system, is presented. The results of the construction of temperature maps of Novosibirsk using this algorithm are discussed.

scholarly journals CHINA FOREST COVER EXTRACTION BASED ON GOOGLE EARTH ENGINE

2020 ◽  
Vol XLII-3/W10 ◽  
pp. 855-862 ◽  
Author(s):  
Y. T. Guo ◽  
X. M. Zhang ◽  
T. F. Long ◽  
W. L. Jiao ◽  
G. J. He ◽  
...  
Keyword(s):  
Random Forest ◽  
Large Scale ◽  
Forest Cover ◽  
Survey Methods ◽  
Coniferous Forest ◽  
Google Earth ◽  
Forest Area ◽  
Landsat 8 ◽  
Forest Change ◽  
Google Earth Engine

Abstract. Forest cover rate is the principal indice to reflect the forest acount of a nation and region. In view of the difficulty of accurately calculating large-scale forest area by traditional statistical survey methods, it is proposed to extract China forest area based on Google Earth Engine platform. Trained by the enough samples selected through the Google Earth software, there are nine different random forest classifiers applicable to their corresponding zones. Using Landsat 8 surface reflectance data of 2018 year and the modified forest partition map, China forest cover is generated on the Google Earth Engine platform. The accuracy of China's forest coverage achieves 89.08%, while the accuracy of Global Forest Change datasets of Maryland university and Japan’s ALOS Forest/Non-Forest forest product reach 87.78% and 84.57%. Besides, the precision of tropical/subtropical forest, temperate coniferous forest as well as nonforest region are 83.25%, 87.94% and 97.83%, higher than those of other’s accuracy. Our results show that by means of the random forest algorithm and enough samples, tropical and subtropical broadleaf forest, temperate coniferous forest and nonforest partition can be extracted more accurately. Through the computation of forest cover, our result shows that China has a area of 220.42 million hectare in 2018.

scholarly journals Coarse-Resolution Satellite Images Overestimate Urbanization Effects on Vegetation Spring Phenology

2020 ◽  
Vol 12 (1) ◽  
pp. 117 ◽  
Author(s):  
Jiaqi Tian ◽  
Xiaolin Zhu ◽  
Jin Wu ◽  
Miaogen Shen ◽  
Jin Chen
Keyword(s):  
Time Series ◽  
Spatial Resolution ◽  
Rural Areas ◽  
Urban Areas ◽  
Satellite Images ◽  
Landsat 8 ◽  
Spring Phenology ◽  
Rural And Urban Areas ◽  
Ndvi Time Series ◽  
The Impact

Numerous investigations of urbanization effects on vegetation spring phenology using satellite images have reached a consensus that vegetation spring phenology in urban areas occurs earlier than in surrounding rural areas. Nevertheless, the magnitude of this rural–urban difference is quite different among these studies, especially for studies over the same areas, which implies large uncertainties. One possible reason is that the satellite images used in these studies have different spatial resolutions from 30 m to 1 km. In this study, we investigated the impact of spatial resolution on the rural–urban difference of vegetation spring phenology using satellite images at different spatial resolutions. To be exact, we first generated a dense 10 m NDVI time series through harmonizing Sentinel-2 and Landsat-8 images by data fusion method, and then resampled the 10 m time series to coarser resolutions from 30 m to 8 km to simulate images at different resolutions. Afterwards, to quantify urbanization effects, vegetation spring phenology at each resolution was extracted by a widely used tool, TIMESAT. Last, we calculated the difference between rural and urban areas using an urban extent map derived from NPP VIIRS nighttime light data. Our results reveal: (1) vegetation spring phenology in urban areas happen earlier than rural areas no matter which spatial resolution from 10 m to 8 km is used, (2) the rural–urban difference in vegetation spring phenology is amplified with spatial resolution, i.e., coarse satellite images overestimate the urbanization effects on vegetation spring phenology, and (3) the underlying reason of this overestimation is that the majority of urban pixels in coarser images have higher diversity in terms of spring phenology dates, which leads to spring phenology detected from coarser NDVI time series earlier than the actual dates. This study indicates that spatial resolution is an important factor that affects the accuracy of the assessment of urbanization effects on vegetation spring phenology. For future studies, we suggest that satellite images with a fine spatial resolution are more appropriate to explore urbanization effects on vegetation spring phenology if vegetation species in urban areas is very diverse.

scholarly journals A Classification of Tidal Flat Wetland Vegetation Combining Phenological Features with Google Earth Engine

2021 ◽  
Vol 13 (3) ◽  
pp. 443
Author(s):  
Nan Wu ◽  
Runhe Shi ◽  
Wei Zhuo ◽  
Chao Zhang ◽  
Bingchan Zhou ◽  
...  
Keyword(s):  
Tidal Flat ◽  
Vegetation Index ◽  
Google Earth ◽  
Wetland Vegetation ◽  
Species Level ◽  
Landsat 8 ◽  
Harmonic Model ◽  
Temporal Features ◽  
Google Earth Engine ◽  
Time Series Images

The composition and distribution of wetland vegetation is critical for ecosystem diversity and sustainable development. However, tidal flat wetland environments are complex, and obtaining effective satellite imagery is challenging due to the high cloud coverage. Moreover, it is difficult to acquire phenological feature data and extract species-level wetland vegetation information by using only spectral data or individual images. To solve these limitations, statistical features, temporal features, and phenological features of multiple Landsat 8 time-series images obtained via the Google Earth Engine (GEE) platform were compared to extract species-level wetland vegetation information from Chongming Island, China. The results indicated that (1) a harmonic model obtained the phenological characteristics of wetland vegetation better than the raw vegetation index (VI) and the Savitzky–Golay (SG) smoothing method; (2) classification based on the combination of the three features provided the highest overall accuracy (85.54%), and the phenological features (represented by the amplitude and phase of the harmonic model) had the greatest impact on the classification; and (3) the classification result from the senescence period was more accurate than that from the green period, but the annual mapping result on all seasons was the most accurate. The method described in this study can be applied to overcome the impacts of the complex environment in tidal flat wetlands and to effectively classify wetland vegetation species using GEE. This study could be used as a reference for the analysis of the phenological features of other areas or vegetation types.

Sign in / Sign up

Export Citation Format

Share Document