Show simple item record

dc.contributorVirginia Tech GIS & Remote Sensing 2014 Research Symposiumen_US
dc.contributor.authorMalambo, Lonesomeen_US
dc.contributor.authorHeatwole, Conrad D.en_US
dc.date.accessioned2014-11-04T19:39:18Z
dc.date.available2014-11-04T19:39:18Z
dc.date.issued2014en_US
dc.identifier.urihttp://hdl.handle.net/10919/50670
dc.description.abstractThe availability of geospatial information has increased dramatically over the years and huge amounts of data are collected each day. Earth observation satellites such as Landsat deliver several hundred gigabytes of data every day, and while free, the sheer sizes of such datasets create computing and storage difficulties even with improved capabilities of current personal computers. Recent developments in cloud computing applications such as the Google Earth Engine® (GEE) can offer solutions to these difficulties by providing an online warehouse of satellite data and superior computing capability needed to analyze these data. With such platforms, the burden of downloading huge amounts of data, storing and processing them can be greatly reduced allowing researchers to concentrate vital data analysis to address complex and expensive environmental challenges being faced today. We present results on Landsat scale mapping of seasonal landscape fires in eastern Zambia using the GEE platform. Our methodology involves automatic training sample selection from burned areas and a subsequent ensemble classification using Random Forests. We mapped fire events using Landsat 8 data – over two districts - from June to October of 2013 with classification accuracies above 90%. Of the 22637 square kilometers mapped, about 60% of it had burned by end of October. The peak of the burning occurred in August with land cover specific fire distribution e.g. in cropland, coinciding with typical land management cycles - which confirms the anthropogenic link of fire in the area.en_US
dc.format.mimetypeapplication/pdfen_US
dc.language.isoen_USen_US
dc.rightsIn Copyright (InC)en_US
dc.rightsThis Item is protected by copyright and/or related rights. Some uses of this Item may be deemed fair and permitted by law even without permission from the rights holder(s). For other uses you need to obtain permission from the rights holder(s).en_US
dc.rights.urihttp://rightsstatements.org/vocab/InC/1.0/
dc.subjectBurned area mappingen_US
dc.subjectForest firesen_US
dc.subjectMulti-temporal analysisen_US
dc.subjectZambiaen_US
dc.subjectGoogle Earth engineen_US
dc.subjectVegetation mappingen_US
dc.titleUsing Cloud Computing for Mapping Seasonal Landscape Fires in Eastern Zambiaen_US
dc.typePosteren_US
dc.rights.holderMalambo, Lonesomeen_US
dc.rights.holderHeatwole, Conrad D.en_US
dc.type.dcmitypeTexten_US
dc.type.dcmitypeImageen_US


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record