Current Journal of Applied Science and Technology, ISSN: 2457-1024; 2231-0843 (old),Vol.: 22, Issue.: 6
Geospatial Modeling and Prediction of Land Use/Cover Dynamics in Onitsha Metropolis, Nigeria: A Sub-pixel Approach
S. U. Onwuka1, P. S. U. Eneche2* and N. A. Ismail2 1Department of Environmental Management, Nnamdi Azikiwe University, Awka, Nigeria. 2Department of Geography and Environmental Studies, Kogi State University, Anyigba, Nigeria.
S. U. Onwuka1, P. S. U. Eneche2* and N. A. Ismail2
1Department of Environmental Management, Nnamdi Azikiwe University, Awka, Nigeria.
2Department of Geography and Environmental Studies, Kogi State University, Anyigba, Nigeria.
(1) Charles W. Recha, Department of Geography, Faculty of Environment and Resources Development, Egerton University, Kenya.
(1) Khairul Nizam Tahar, Universiti Teknologi MARA, Malaysia.
(2) Jesus Soria-Ruiz, Geomatics Lab. National Institute of Research for Forestry Agricultural and Livestock (INIFAP), Mexico.
Complete Peer review History: http://www.sciencedomain.org/review-history/20325
Based on a sub-pixel approach, this study analysed the Land Use/Cover (LU/C) dynamics of Onitsha Metropolis in Anambra State, Nigeria. Landsat TM/ETM+ satellite imageries of 1986, 2001 and 2016 were characterized into different LU/Cs using Ridd’s Vegetation, Impervious Surface, Soil and Water (VIS-W) model via Linear Spectral Mixture Analysis (LSMA). LU/C endmember fractions obtained were hardened to produce the final LU/C maps of the study area, per study years considered. Cellular Automata Markov (Ca-Markov) chain and the Land Change Modeler (LCM) were used to predict future LU/C for the year 2031 and the transition of each LU/C categories between 2016 and 2031, respectively. Also, the Chi-square test was used to test the significance of change in LU/C fractions between 2016 and 2031. ArcGIS 10.5, Idrisi Selva and Statistical Package for Social Science (SPSS 22) were used to perform the analyses. The result of the LU/C classification on one hand, revealed the dynamics of LU/C endmember fractions for the study years and on the other hand, revealed the actual area coverage of each LU/C category. It showed clearly that vegetation reduced drastically over the three epochs from 178.72sq.km in 1986 to 147.70 sq.km in 2001 and slightly to 140.87 sq.km in 2016; impervious surface increased from 26.10 sq.km in 1986 to 62.28 sq.km in 2016; soil cover decreased from 8.65 sq.km in 1986 to 3.10 sq.km in 2016; and water cover, experienced an increase from 11.44 sq.km in 1986 to 18.75 sq.km in 2016. The Ca-Markov and the LCM models further revealed that all LU/C fractions, apart from soil possessed very high probability of being retained in 2031, thus, are envisaged to be slightly modified in future. However, the result of the Chi-square test confirms no statistically significant difference in the LU/C fractions between 2016 and 2031 (P=.964, α = .05). Therefore, it was upheld in this study that the rapidity of urbanisation in Onitsha Metropolis has drastically reduced while the degree or intensity of urbaneness was on the increase, especially in recent years and the same trend is expected to continue except otherwise, other factors set-in. The continuum-based approach of this study however, presents an objective means of characterizing LU/C fractions and recommended in modelling the urban fabrics of any area, especially when other non-linear and chaotic urban phenomena, such as Urban Heat Island (UHI); urban land suitability/compatibility, flooding, physiological discomfort, etc. are of interest.
Linear spectral mixture analysis; cellular automata Markov chain; land change modeler; endmember fraction.
Full Article - PDF Page 1-18
DOI : 10.9734/CJAST/2017/35294Review History Comments