Naslov
Land use land cover change analysis using Landsat imagery in Međimurje county, Croatia

Autori
Horvat, Zlatko

Vrsta, podvrsta i kategorija rada
Sažeci sa skupova, sažetak, znanstveni

Skup
GI_Forum 2013 - Creating the GISociety

Mjesto i datum
Salzburg, Austrija, 02.06.2013. - 05.06.2013

Vrsta sudjelovanja
Poster

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
Land Use; Land Cover; Change Detection; Remote Sensing; Landsat Satellite Imagery; Međimurje County; Hybrid Classification

Sažetak
Timely and precise information about Land Use Land Cover (LULC) change detection of the Earth’s surface is extremely important for understanding relationships and interactions between human and natural phenomena for better management of decision making. It is well known that there are only a few places on the Earth that are still in their natural state and that have not been affected by human activity in some way. These human activities result in significant land use changes at regional and local scales together with ecological, socio-economic and aesthetical impacts. This study evaluates LULC changes in Međimurje County, Croatia, between 1978, 1992 and 2007 using Landsat satellite images. Spatial dynamics of LULC changes were quantified using three Landsat satellite images: MSS, TM and ETM+. Three series of the LULC maps of 1978, 1992 and 2007 were produced. A post classification technique was used based on a hybrid classification approach comprised of unsupervised classification based on “hill climbing” cluster algorithm, supervised classification based on the Maximum Likelihood Classifier (MLC) and introduction of human-knowledge from field experience. Careful manual delineation on the topographic maps 1:25, 000 and high aerial photographs were done around all urban areas of Međimurje County. The urban areas were clipped out of the image data later and classified separately from other data with an unsupervised classification. Based on the Anderson classification system, the land-use and land-covers are classified as: water bodies, forest land, barren land, agricultural land and urban built up land. Error matrices were used to assess classification accuracy. The overall accuracies, user’s and producer’s accuracies and the Kappa statistics were derived from the error matrices. The overall accuracies for 1978, 1992 and 2007 were, respectively, 87.67%, 88.96% and 90.84%. Kappa statistics were 80%, 84% and 83%. User's and producer's accuracies of the individual classes were relatively high, ranging from 74% to 94% which indicates a good agreement between the thematic maps generated from images and the reference data. The results of the analysis showed that from 1978 to 2007 the urban area increased approximately 53%, while the agricultural land decreased 12%. Water, forest and barren land also increased 14%, 5% and 49% respectively. The use of the traditional per-pixel based classification in combination with a separate classification of urban areas approach, i.e. with dividing barren agricultural fields from built-up areas, a good knowledge of the research area, GIS tools, availability of the quality ancillary data (aerial photographs and topographic maps) for the selection of training samples for „ground truth“ information required for classification, and for accuracy assessment of achieved classified LU/LC maps, demonstrate the potential of multi-temporal Landsat satellite imagery to provide accurate LULC maps and to analyze changes in LULC that can be used as inputs to land management and policy decisions, to the development of sustainable urban land use planning decisions and for governmental or non-governmental organizations in a variety of purposes as up-to-date information.

Izvorni jezik
Engleski

Znanstvena područja
Geologija

POVEZANOST RADA