rgee: An R package for interacting with Google Earth Engine

Aybar, César; Wu, Qiusheng; Bautista, Lesly; Yali, Roy; Barja, Antony

doi:10.21105/joss.02272

Cited by 107 publications

(97 citation statements)

References 2 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Google Earth Engine (GEE) is a cloud-based platform that facilitates access to high-performance computing resources to process large volumes of geospatial information without being limited to the characteristics of a local machine [40]. GEE processing services were accessed via R Studio Environment using the R package rgee [41] to collect and pre-process information for factors relating to the probability of occurrence. Some base information layers were downloaded from external sources (weather and roads), but these were also incorporated into the GEE platform to unify the location and availability of the information layers.…”

Section: Preprocessingmentioning

confidence: 99%

Open Data and Machine Learning to Model the Occurrence of Fire in the Ecoregion of “Llanos Colombo–Venezolanos”

Barreto

Armenteras

2020

Remote Sensing

View full text Add to dashboard Cite

A fire probability map is an important tool for landscape management, providing better identification of areas prone to fires and helping optimize the allocation of limited resources for fire prevention, control, and management. In this study, the random forest machine learning algorithm was applied to model the probability of fire occurrence in the Colombian-Venezuelan plains (llanos) ecoregion in South America. Information on burned areas was collected using Moderate Resolution Imaging Spectroradiometer (MODIS) Product MCD64A1 for the period 2015–2019. We also used spatial information of related factors that were grouped into four levels of information: topography, human presence, vegetation, and climate-related variables. The model had an accuracy of 94%, which indicates the performance of the model was excellent. The cartography generated from the model can be used as base information in the context of fire management in the region, to identify areas for prioritizing efforts and attention. The probability of occurrence zoning results indicates that the very low category covers the largest area (28.2%), followed by low (23.2%), very high (17.6%), moderate (17.2%), and high (13.8%).

show abstract

Section: Preprocessingmentioning

confidence: 99%

Open Data and Machine Learning to Model the Occurrence of Fire in the Ecoregion of “Llanos Colombo–Venezolanos”

Barreto

Armenteras

2020

Remote Sensing

View full text Add to dashboard Cite

show abstract

“…The annotation of fractional units can be more efficient when using online platforms e.g., Geo-Wiki (Tsendbazar et al 2018). Moreover, the advantages of using GEE for data access and preprocessing together with the statistical functions of R can be both utilized using the new rgee package (Aybar et al 2020). The fractional cover approach can also be applied for baseline and large-scale mapping of other forest trees with economic potential such as Almaciga (Agathis philippinensis Warb., Monsunia), Benguet pine (Pinus kesiya Royle ex Gordon var.…”

Section: Future Directions Of Bamboo Resources Mappingmentioning

confidence: 99%

Towards nationwide mapping of bamboo resources in the Philippines: testing the pixel-based and fractional cover approaches

Dida

Araza

Eduarte

et al. 2021

International Journal of Remote Sensing

View full text Add to dashboard Cite

In tropical and subtropical countries, the awareness on the importance of bamboos to the environment and economy is increasing and so is the demand for spatial bamboo information. However, mapping bamboos especially those naturally grown has been challenging, as these grasses are often mixed with other land-use and land-cover (LULC). In this study, we used Sentinel 1 and Sentinel 2 remote sensing (RS) images, and their vegetation indices to accurately map the bamboos of Iloilo province in the Philippines using:(1) pixel-based method that mapped bamboos and other LULC at 10 m resolution, and (2) fractional cover method that mapped bamboo cover at 100 m resolution (% ha −1 ). A random forest model was trained for each method and then validated per hectare basis using a 50:50 training-validation ratio of a stratified random sample. The fractional cover method showed 0.34 higher Nash-Sutcliffe Efficiency (NSE) and 5.10% lower Root Mean Square Error (RMSE) than the pixel-based method. Further validation within upland and lowland sites also favoured the fractional cover method, but the results of the two methods were closer in the upland site (bamboo plantation). Errors at 10 m resolution especially in the lowlands were mostly commission errors, likely because of the spectral similarity and proximity between bamboos and > 14 vegetations. Averaging the RS inputs into 100 m resulted in at most 12% separation of reflectance values among bamboos, forests, and other vegetations. Using the bamboo cover map, a total of 14,795 (± 1,283) ha bamboos and 7.45 (± 4.20) million harvestable culms (poles) were estimated for the whole province, where 54% come from the lowland. We suggest using the fractional cover method for nationwide baselining of bamboo resources.

show abstract

“…All our climate hypotheses are based on combinations of temperature and precipitation (see Table 1 for full list and descriptions). Thus, we used the Google Earth Engine (GEE) data catalogue "PRISM Daily Spatial Climate Dataset AN81d" (Daly et al, 2015) to extract climate data using the R package rgee (Aybar et al, 2020). This dataset includes daily measures of temperature (°C) and total daily precipitation (mm; rain and melted snow) for the USA, assimilated from many weather stations across the country and interpolated to create a smooth raster.…”

Section: Predictor Variablesmentioning

confidence: 99%

Predicted climate‐induced reductions in scavenging in eastern North America

Marneweck

Katzner

Jachowski

2021

Global Change Biology

View full text Add to dashboard Cite

Scavenging is an important function within ecosystems where scavengers remove organic matter, reduce disease, stabilize food webs, and generally make ecosystems more resilient to environmental changes. Global change (i.e., changing climate and increasing human impact) is currently influencing scavenger communities. Thus, understanding what promotes species richness in scavenger communities can help prioritize management actions. Using a long‐term dataset from camera traps deployed with animal carcasses as bait along a 1881 km latitudinal gradient in the Appalachian Mountains of eastern USA, we investigated the relative impact of climate and humans on the species richness and diversity of vertebrate scavengers. Our most supported models for both mammalian and avian scavengers included climatic, but not human, variables. The richness of mammalian and avian scavengers detected was highest during relatively warm (5–10°C) and dry (100–150 mm precipitation) winters, when food was likely limited and both reliance on and detection of carrion was high. The diversity of mammalian and avian scavengers detected was highest under drier conditions. We then used these results to project the future species richness of scavengers that would be detected within our sampling area and under the climate scenario of 2070 (emissions level RCP8.5). Our predictions suggest up to 80% and 67% reductions, respectively, in the richness of avian and mammalian scavengers that would be detected at baited sites. Climate‐induced shifts in behavior (i.e., reduction in scavenging, even if present) at this scale could have cascading implications for ecosystem function, resilience, and human health. Further, our study highlights the importance of conducting studies of scavenger community dynamics within ecosystems across wide spatial gradients within temperate environments. More broadly, these findings build upon our understanding of the impacts of climate‐induced adjustments in behavior that can likely have negative impacts on systems at a large scale.

show abstract

rgee: An R package for interacting with Google Earth Engine

Cited by 107 publications

References 2 publications

Open Data and Machine Learning to Model the Occurrence of Fire in the Ecoregion of “Llanos Colombo–Venezolanos”

Open Data and Machine Learning to Model the Occurrence of Fire in the Ecoregion of “Llanos Colombo–Venezolanos”

Towards nationwide mapping of bamboo resources in the Philippines: testing the pixel-based and fractional cover approaches

Predicted climate‐induced reductions in scavenging in eastern North America

Contact Info

Product

Resources

About