Burnt area detection using two post fire images

Remote sensing from satellite is  a good tool for both fire detection and burn area assessment. A multi-temporal approach relies on pre and post fire multispectral  images to derive vegetation maps (e.g. Normalised Differential Vegetation Index-NDVI, dNBR . We have   considered two test cases:  one in UK were only post fire cloud free Sentinel 2 data are available and one in Italy were pre -fire, active- fire and post- fire Sentinel 2 A data are available. 

The Dovestones- Stalybridge  wildfire occurred on 10th  April 2016.  Data acquired by Sentinel 2A on 20 April 2016 and 19 July 2016 were used  to derive  an "a posteriori" dNDVI -differential NDVI- by using two post images. A validation fieldwork  was conducted  on July 2016 to identify  re-growing and burn patches  to  validate the obtained results.

The Italian test site is located in S. Marinella area Latium Italy, images acquired by Sentinel 2A pre fire on 11 May  2017 and two post 9 August 2017 and 27 November 2017 . The results are promising and the techniques can be used to  validate burn area detection retrieved by using new SAR based methods or to complement more traditional methods based on dNDVI or dNBR by using a pre- post pairs. Work is  in progress to derive an optimized threshold approach to be tested on a range of test sites. .