Geografie 2022, 127, 219-240

Looking into the green roof scenario to mitigate flash flood effects in Mamak, Turkey, via classifying images of Sentinel-1, 2, and PlanetScope satellites with LibSVM algorithm in Google Earth Engine cloud platform

Sima Pouya1ID, Majid Aghlmand2ID, Fevzi Karsli3ID

1Inonu University, Faculty of Fine Arts and Design, Department of Landscape Architecture, Malatya, Turkey
2Eskisehir Technical University, Civil Engineering Department, Eskisehir, Turkey
3Karadeniz Technical University, Department of Geomatics, Trabzon, Turkey

Received November 2021
Accepted May 2022

This research aimed to increase the green space factor to mitigate flash flood effects on urban storm water runoff in the Ankara Mamak region and to minimize the damages by flash floods. The land use/cover map was first obtained by using the images of Sentinel-1, Sentinel-2, and PlanetScope satellites with the LIBSVM algorithm on the Google Earth Engine. The GSF value was then calculated and it was low (0.26) compared to world standards. This study was proposed as a solution for the flood disaster, using the extensive green roof scenario. After green roof conversion scenarios, the GSF value was recalculated. It was found to be above the minimum of green infrastructure that human settlements should achieve, regardless of density or land use (0.43). Offering high resolution images and the possibility of processing them via different algorithms of machine learning has revolutionized the environmental and urban-related studies as they help urban managers and planners to make decisions accurately and quickly.


58 live references