Prediction of built-up changes and urban growth using remotely sensed data

The world is rapidly moving towards urbanization, and with large populations living in cities, and ever increasing population in urban areas, urban sprawl has occurred in many cities around the world. Lack of urban planning and management regarding the development of urban sprawl has been known as the source of many problems in cities around the world. Urban sprawl negatively impacts the environment, quality of life, social equity, climate change, air pollution and LST. Therefore, one of the goals of this study is to show the pattern of urban growth from the past to present, and predict the future of urban growth that may occur. It can be considered as an innovation of this study, which was less prominent in previous studies.

The study area is Rasht, where its natural attractions and tourism characteristics have increased its population and physical development. This research utilized Landsat 5 (TM), 7 (ETM+), and 8 (OLI/TIRS) images. In the first step, the pre-processing operations including geometric correction, atmospheric correction, and radiometric correction were performed on the remote sensing images. In the next step, Normalized Difference Impervious Index (NDISI) values were computed and employed to extract the impervious surface information in urban area. Then a hybrid cellular automaton–Markov (CA-Markov) model was used to predict both the quantity and spatial distribution of impervious surfaces in the city of Rasht. The performances of these methods were evaluated using 300 randomly selected samples. Finally, statistical analysis has been used to show the growth pattern of Rasht from its past to present and also to predict the future.

The results of this study indicated that the impervious surface of this city can be extracted with high accuracy (from 86.12 percent to 89.88 percent) using NDISI computed from Landsat images. Moreover, the accuracy of the CA-Markov model through prediction of impervious surface for the year 2018 was about 83.21 percent. With regard to the results of this work, the observed and expected urban growth results are not consistent with each other. The analysis of degree-of-freedom ( ) and Shannon's entropy ( ) are reflecting the urban sprawl pattern. Then, H and were used to compute the degree of goodness ( ). This parameter demonstrated that the growth pattern of Rasht was inappropriate.

This study shows that the sprawl of the city of Rasht can be characterized by a scattered growth that is expected to be worse in future, if management measures are not taken by governmental authorities. The results of this study can be useful for future urban planning and decision making such as preventing vertical land use changes.

* The formula is not displayed correctly.