Διερεύνηση δυνατοτήτων δορυφορικής τηλεπισκόπησης στην Κύπρο για την παρακολούθηση περιοχών μετά το φαινόμενο πυρκαγιάς
Date Issued
2015
Author(s)
Advisor
Abstract
It is indisputably constitutes that forest is renewable resource and provides unlimited capabilities as in environmental environmental, cultural, social and economic development of a region. Wildfires constitutes an integral part of an ecosystem which brings disastrous consequences affecting both the vegetation and the structure and function of the ecosystem.
Satellites data provide the advantage of monitoring and repeatable coverage of the earth surface, so Satellite Remote Sensing widely use in all phases of the management of forest fires. More specifically, it is used both for prevention, detection and direct mapping, and for estimate short and long term arising effects.
In this thesis it was used for initial detect of burned areas, time series of satellite data Landsat 5 TM and 7 ETM +. The forest of Macheras and Saittas in Cyprus were the selected study areas, where there was devastating wildfire in 2001 and 2007 respectively. At first, for detection and mapping of the burnt area, was created images by band combination (R, G, B) of the satellite images before and after wildfires, from which only the best band combination was selected.
Then, Normalized Difference Vegetation Index was calculated (NDVI) for both burnt areas and healthy vegetation. The index values were used to create diagrams, which proved that the vegetation recovery rate in Saitta was higher than the one of Macheras. Besides, there were calculated other spectral vegetation indices (DVI, RDVI, TVI, SAVI, OSAVI, GEMI, IPVI) and then mapped for better visual of the results.
For assessing the gravity of the burnt area was calculated the index Normalized Burn Ratio (NBR) on satellite images before and after the fire, and then created a map of the dNBR index. Finally, applied supervised classification of the area Macheras in time series of satellite images, with the ultimate purpose of illustrating the gradual development of vegetation.
Satellites data provide the advantage of monitoring and repeatable coverage of the earth surface, so Satellite Remote Sensing widely use in all phases of the management of forest fires. More specifically, it is used both for prevention, detection and direct mapping, and for estimate short and long term arising effects.
In this thesis it was used for initial detect of burned areas, time series of satellite data Landsat 5 TM and 7 ETM +. The forest of Macheras and Saittas in Cyprus were the selected study areas, where there was devastating wildfire in 2001 and 2007 respectively. At first, for detection and mapping of the burnt area, was created images by band combination (R, G, B) of the satellite images before and after wildfires, from which only the best band combination was selected.
Then, Normalized Difference Vegetation Index was calculated (NDVI) for both burnt areas and healthy vegetation. The index values were used to create diagrams, which proved that the vegetation recovery rate in Saitta was higher than the one of Macheras. Besides, there were calculated other spectral vegetation indices (DVI, RDVI, TVI, SAVI, OSAVI, GEMI, IPVI) and then mapped for better visual of the results.
For assessing the gravity of the burnt area was calculated the index Normalized Burn Ratio (NBR) on satellite images before and after the fire, and then created a map of the dNBR index. Finally, applied supervised classification of the area Macheras in time series of satellite images, with the ultimate purpose of illustrating the gradual development of vegetation.
File(s)![Thumbnail Image]()
Name
Περίληψη.pdf
Size
176.88 KB
Format
Adobe PDF
Checksum (MD5)
4bdc479b49d3e5ef808d154050ee4fdf