Details

Autor(en) / Beteiligte

• da Silva, Fábio Miguel Domingues

Titel

Lulc Map Comparison: Comparison and Validation of Land Use Land Cover Maps Derived from Satellite Imagery

Ort / Verlag

ProQuest Dissertations & Theses

Erscheinungsjahr

2022

Quelle

ProQuest Dissertations & Theses A&I

Beschreibungen/Notizen

• The technological evolution of remote sensing techniques has allowed for the ever-growing creation of land use land cover maps. Nowadays mapping entities from all around the globe are capable of producing their own products, be it for their region or country, or for a whole continent or even the whole world. However, this raises an issue regarding the comparison of these maps, in one sense direct comparison is difficult due to the lack of harmonization of the data, be it the nomenclature’s classification, or technical specifications of the imagery. In another sense, the creation of global cover maps, or continent-wide maps, hinges the ability to accurately classify LULC appropriately due to the complexity of land cover, often leaving specific regions of the map with a less accurate classification, even though the overall one is good. Throughout this study, five maps from five different mapping entities will be compared and evaluated, these maps are COSsim, ELC10, ESA Worldcover, ESRILC and S2GLC. The study area is Continental Portugal, and the main objective is to understand how the international mapping entities’ maps compare with the Portuguese map of COSsim, by observing nomenclature differences and accuracy scores. As well as understand what the impact in accuracy is, in European cover or World cover maps, when only analyzing them for the study area of Continental Portugal. The results obtained showed that most international maps proved to have a much smaller accuracy score for Continental Portugal, most of these even having a 20% to 30% drop in their overall accuracy. This research helps understand the necessity for the harmonization of nomenclatures, and at the same time investment necessary for national mapping entities to create their own more accurate maps.