Приказ основних података о документу

Automated mapping of climatic variables using spatio-temporal geostatistical methods

dc.contributor.advisorBajat, Branislav
dc.contributor.otherDucić, Vladan
dc.contributor.otherNestorov, Ivan
dc.contributor.otherTošić, Ivana
dc.contributor.otherHengl, Tomislav
dc.creatorKilibarda, Milan
dc.date.accessioned2016-01-05T11:52:02Z
dc.date.accessioned2019-05-01T00:04:33Z
dc.date.available2016-01-05T11:52:02Z
dc.date.available2019-05-01T00:04:33Z
dc.date.issued2013
dc.identifier.urihttp://eteze.bg.ac.rs/application/showtheses?thesesId=1380
dc.identifier.urihttps://fedorabg.bg.ac.rs/fedora/get/o:8525/bdef:Content/download
dc.identifier.urihttp://vbs.rs/scripts/cobiss?command=DISPLAY&base=70036&RID=513383570
dc.identifier.urihttp://nardus.mpn.gov.rs/123456789/2196
dc.identifier.urihttps://grafar.grf.bg.ac.rs/handle/123456789/1670
dc.description.abstractJavno dostupni meteorološki podaci, kako sa stanica tako i iz daljinske detekcije, korišćeni su za prostorno vremensku interpolaciju temperature vazduha iznad površine Zemlje. Zastupljenost i pogodnost javno dostupnih podataka je ocenjena, kroz tri aspekta kontrole kvaliteta: (a) zastupljenost u geografskom i prostornom domenu, (b) zastupljenost u karaktestičnom prostoru (feature space; bazirano na MaxEnt metodi), kao i (c) pogodnost korišćenja podataka za prostorno-vremensku predikciju (na osnovu kros-validacije prostorno-vremnskog regresionog kriginga). Rezultati pokazuju da je kombinovani set podataka (GSOD i ECA&D) značajno klasteriran i u geografskom i u karakterističnom prostoru. Uprkos klasteriranju, preliminarni rezultati globalne interpolacije primenom prostorno-vremenskog regresionog kriginga koristeći merenja sa stanica i snimke daljinske detekcije su pokazali da se tako mogu dobiti precizne globalne karte dnevne temperature. Oko 9000 stanica kombinovanog seta podataka (GSOD i ECA&D) je korišćeno za prostorno-vremensko geostatističko modeliranje i predikciju dnevnih temperatura u rezoluciji 1 km, iznad površine Zemlje. Za predikciju srednjih, minimalnih i maksimalnih temperatura korišćen je regresioni kriging uz pomoćne prediktore: MODIS LST 8-dnevni snimci, topografski lejeri (DEM i TWI) i geometrijski temperaturni trend. Model i predikcija se odnose na 2011 godinu, ali ista metodologija bi se mogla primeniti od 2001 godine do danas (od kada su dostupni MODIS snimci). Rezultati pokazuju da je prosečna tačnost predikcije za srednju, minimalnu i maksimalnu temperaturu vazduha oko ±2°C za oblastigusto pokrivene stanicama i između ±2°C i ±4°C za oblasti koje su slabo pokrivene stanicama. Najniža tačnost predikcije je dobijena u planinskim predelima i na Antartiku, oko 6°C. R softverski paket, meteo, je razvijen kao resenje za automatsko kartiranje. Razvijen je i paket plotGoogleMaps za automatsku vizuelizaciju na Web-u, koristeći Google Maps API.sr
dc.description.abstractPublicly available global meteorological data sets, from ground stations and remote sensing, are used for spatio-temporal interpolation of air temperature data for global land areas. Publicly available data sets were assessed for representation and usability for global spatio-temporal analysis. Three aspects of data quality were considered: (a) representation in the geographical and temporal domains, (b) representation in the feature space (based on the MaxEnt method), and (c) usability i.e. fitness of use for spatio-temporal interpolation (based on cross-validation of spatio-temporal regression-kriging models). The results show that clustering of meteorological stations in the combined data set (GSOD and ECA&D) is significant in both geographical and feature space. Despite the geographical and feature space clustering, preliminary tested global spatio-temporal model using station observations and remote sensing images, shows this method can be used for accurate mapping of daily temperature. Around 9000 stations from merged GSOD and ECA&D daily meteorological data sets were used to build spatio-temporal geostatistical models and predict daily air temperature at ground resolution of 1 km for the global land mass. Predictions were made for the mean, maximum and minimum temperature using spatio-temporal regression-kriging with a time series of MODIS 8 day images, topographic layers (DEM and TWI) and a geometrical temperature trend as covariates. The model and predictions were built for the year 2011 only, but the same methodology can be extended for the whole range of the MODIS LST images (2001–today). The results show that the average accuracy for predicting mean, maximum and minimum daily temperatures is RMSE = ± 2°C for areas densely covered with stations, and between ± 2°C and ± 4°C for areas with lower station density. The lowest prediction accuracy was observed in highlands (> 1000 m) and in Antarctica with a RMSE around 6°C. Automated mapping framework is developed and implemented as R package meteo. Likewise, packageplotGoogleMaps for automated visualisation on the Web, base on Google Maps API isdeveloped.en
dc.formatapplication/pdf
dc.languageen
dc.publisherУниверзитет у Београду, Грађевински факултетsr
dc.rightsopenAccess
dc.rights.urihttps://creativecommons.org/licenses/by-nc-nd/4.0/
dc.sourceУниверзитет у Београдуsr
dc.subjectprostorno-vremenska interpolacijasr
dc.subjectspatio-temporal interpolationen
dc.subjectspatio-temporal krigingen
dc.subjectspace-time variogramen
dc.subjectlinear regressionen
dc.subjectMaxEnten
dc.subjectdaily air temperatureen
dc.subjectMODIS LSTen
dc.subjectglobal modelen
dc.subjectprostorno-vremenski krigingsr
dc.subjectprostornovremenskivariogramsr
dc.subjectlinearna regresijasr
dc.subjectMaxEntsr
dc.subjectdnevne temperature vazduhasr
dc.subjectMODISLSTsr
dc.subjectglobalni modelsr
dc.titleAutomatsko kartriranje klimatskih varijabli primenom prostorno-vremenskih geostatičkih metodasr
dc.titleAutomated mapping of climatic variables using spatio-temporal geostatistical methodsen
dc.typedoctoralThesisen
dc.rights.licenseBY-NC-ND
dc.identifier.fulltexthttps://grafar.grf.bg.ac.rs//bitstream/id/2610/Disertacija.pdf
dc.identifier.rcubhttps://hdl.handle.net/21.15107/rcub_nardus_2196
dc.type.versionpublishedVersion


Документи

Thumbnail

Овај документ се појављује у следећим колекцијама

Приказ основних података о документу