University of Twente Student Theses


Uncertainty modeling for asynchronous time series data with incorporation of spatial variation for Land Use/Land Cover Change

Choudhari, Deepak Kumar (2013) Uncertainty modeling for asynchronous time series data with incorporation of spatial variation for Land Use/Land Cover Change.

[img] PDF
Abstract:Land is vital to the survival of all life on Earth and it is important that we understand the various changes that take place on it. Environmental and anthropogenic drivers are constantly changing the face of the Earth and it is essential that we understand what these drivers are and the change they bring about to the Land use/Land cover. Land Use Land Cover (LULC) change analysis is very important for environmental management purposes as it helps the decision maker in planning for future changes that may occur in that area and it also helps the decision maker realize the effects of these changes on humans and their environment. Decision makers also have to identify, what are the factors that affect the LULC change (such as population, agricultural growth etc.). Therefore the quality of the LULC change map is essential for making more accurate decisions. In this project, uncertainty modeling has been performed in Land use/cover change. Uncertainty modeling quantifies the variation in results obtained by a modeling process for decision-making. Remote Sensing and Geographical Information Systems apply different types of operations (e.g. classification, rescaling) on spatial datasets to produce maps or to extract spatial information from the dataset; there is always some uncertainty in these operations. Uncertainty considers aspects like error and incompleteness of the input data as well as the output data. This project uses a CA-Markov hybridized approach to model LULC change in the Upper Ganaga Basin. LULC change is detected by using different time period images and various driver datasets (e.g. soil moisture, temperature, and slope). There is some amount of uncertainty in the available observation, spatial distribution and resolution of the datasets. We have to model the redistribution and aggregation uncertainty for the driver data in order to come up with an outcome that has minimum error. Spatial aggregation is used in conversion of LULC datasets to thematic raster datasets. The spatial aggregation methods used for this project are Random role-based and Major role-based The results show that major rule-based aggregations proved to be more accurate than random rule-based aggregations for the Upper Ganga basin. The highest percentage of change was observed in the classes; snow and ice and barren land. Keywords: LULC Change, CA-Markov, Major Rule Based, Random Rule Based
Item Type:Essay (Master)
Faculty:ITC: Faculty of Geo-information Science and Earth Observation
Programme:Geoinformation Science and Earth Observation MSc (75014)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page