Land use change detection of Ardabil province in last two decades using visual and digital interpretation of sattelite imagery

Kakehmami, Azad (2016) Land use change detection of Ardabil province in last two decades using visual and digital interpretation of sattelite imagery. Masters thesis, University of Mohaghegh Ardabili.

Text (بررسي تغييرات کاربری اراضي استان اردبیل در دو دهه اخير به دو صورت تفسیر چشمی و رقومی با استفاده از تصاویر ماهواره‌ای) Azad Kakehmami.pdf Download (1MB)

Abstract

With the evaluation of change detection, the possibility of planning and proper management to reduce the effect of land degradation will be possible. Provding landuse map with high accuracy and suitable quality is one of the fundamental requirments for plannig in different disiplines, specially in natural resources including range management. The aim of this study was to produce land use map of Ardabil province by visual interpretation using Google Earth’s imagery, and digitizing image proceesing using Landsat Thematic Mapper (1987) and Operational Land Imager (2015) imagary and change detection in the defined period. Initialy, the efficiency of Visual Interpretation, Maximum Likelihood, Artificial Nural Netwok (ANN) and Object based methods were compared and after evaluation of the accuracy of each produced map using ground conntrol points, the accuracy of them for 1987 image were calculated as 61.3, 60.5 and 82.3 % (visual interpretation was not conducted for this image), and for 2015 image the accuracy were 99.4, 60.8, 94 and 82%, respectively. Maps resultaed from object based method for 1987 and visual interpretation of Google Earth’ imagery for 2015 according to their high accuracy were selected for change detection stage. The results indicated that between 1987 and 2015 rangelands and forests have redused by 75611 and 9412 ha and irigated farming and dry farming and residential areas have increased by 14263, 54097 and 15050 ha. The main changed land use from 1987 to 2015 are rangelands, which have changed to dry farming and irigated farming lands by 141794 and 49836 ha, respectively. Thus, it is strongly required to review management polices and practices. For this pourpoe, satelite data are useful for the identification of land use pattern in a large areas and detection of change throuth the time. Keywords: Ardabil, Land use, Remote sensing, Object Based.

Actions (login required)

View Item