مدل سازی توسعه فیزیکی شهر با ترکیب قابلیت‌های گوگل ارث انجین (Gee) و شبکه عصبی مصنوعی (Mlp) مطالعه موردی: شهر تبریز

امروزه ارزیابی توسعه فیزیکی شهری با استفاده از تکنیک‌های نوین سنجش از دور می‌تواند اطلاعات پایه ای را در اختیار برنامه ریزان قرار دهد و از این طریق نقشی موثر در مدیریت و بهبود کاربری اراضی شهر ایفا کند. هدف از این پژوهش، پایش و ارزیابی توسعه فیزیکی شهر تبریز در دوره 42 ساله (20141972) با استفاده از سامانه جدید (gee) google earth engine و پیش بینی تغییرات گسترش فیزیکی شهر تبریز با استفاده از مدل شبکه عصبی (mlp) است. نتایج این پژوهش نشان از قابلیت بالای فناوری gee در استخراج پهنه‌های شهری طی دوره‌های مختلف دارد، بطوریکه این فناوری بخوبی توانست توسعه فیزیکی شهر تبریز را طی دوره 40 ساله ارزیابی کند. نتایج پیش‌بینی تغییرات حاصل از مدل mlp نشان دهنده‌ی این است که توسعه فیزیکی شهر تبریز در آینده رو به شمال شرقی و جنوب شرقی است و مدل اجرا شده از سال 1975 تا 2014 در قالب gee و برای بیست سال آتی با استفاده از مدل تجربی شبکه عصبی مبتنی بر پرسپترون چندلایه اقدام به شبیه سازی و مدلسازی روند آتی توسعه کلانشهر تبریز نمود.

Modeling physical development by combining the capabilities of Google Earth Engine (GEE) and Artificial Neural Network (MLP) the Case Study: Tabriz

IntroductionUrbanization as a revolution in human culture has transformed human interactions with one another. As the urbanization population grows, the use of the environment is intensified. Studies have shown that increasing population and expanding urbanization are turning urban green spaces into rough and impermeable concrete surfaces, and this trend is especially serious in developing countries and the Third World. Since urban growth is a complex phenomenon in which a number of variables interact nonlinearly, the use of ANNs to model urban development and growth is perfectly reasonable. Artificial neural networks with nonlinear mapping structure have been developed for modeling interconnected systems such as the brain consisting of neurons. The artificial neural network is independent of the statistical distribution of data and does not require any specific statistical variables, so this feature facilitates the combination of remote sensing data and GIS. Currently, remote sensing science is changing a fundamental paradigm in which one or twoimage interpretation approaches pave the way for a wide array of datarich applications. These improvements are facilitated by the GEE Satellite Image Processing System. The purpose of this research is to introduce a new system (GEE), to investigate and analyze this web portal, its application in monitoring and evaluation of human habitat changes (GHSL) and to map the relationship created using MLP model to predict physical development changes in Tabriz.Materials and MethodsIn this study, the Google Earth Engine (GEE) satellite image processing online system was used to process and extract the global GHSL product, and then the MLP model of Terset was used to predict changes.Results and Discussion In this study, it was attempted to analyze and analyze Landsat satellite images in a few minutes in order to prepare physical development map of Tabriz city without using hard data and to predict future development changes using the data available in Google Inheritance Satellite Image Processing System. Physically measure the city using the MLP model. GEE online processor has been able to map the growth of urbanization in the Tabriz city over the past six years. With the increase in urbanization over the past 40 years in the city of Tabriz, we have seen the destruction of about 38% of gardens and agriculture in the city, and even this system of rapid population growth in recent years (2014) on the outskirts of Tabriz as the main center of recent earthquakes.Conclusion It has shown the city of Tabriz and is also witnessing a growing trend towards physical development of the city in this part of Tabriz. The results of the MLP model show that the physical development of Tabriz in the future is northeastward and on the outskirts of Mount Aoun bin Ali.