Evaluation of Several Spectral Indices to Produce Post Fire Maps (Case Study: Cedar-Fir Protected Area, Syria)

Ola Merhej (1*)

(1). General Organization of Remote Sensing, Latakia, Syria.

(*Corresponding author: Ola Merhej. E-Mail: olamerhej@gmail.com).

Received: 21/05/2021                Accepted:19/02/2022

Abstract: 

In this study, the application of four spectral indices (BAIS2, GEMI, NDVI, NBR) derived from sentinel 2 to map the burned areas for the Cedar-Fir Protected area was evaluated. Although the study area is considered as natural reserve, which is located in the top of coastal mountains (Slinfeh) in Syria, it was exposed to a large fire in the Fall of 2020. We used two Sentinel 2A images acquired before and after the fire on September 4 and October 22 2020. The studied Post-fire indices showed a similar spatial distribution for the burned area, with a preference for BIAS2 index. BIAS2 and dBAIS2 gave the best coefficient of variation value, which ​​reached 40%. On the other hand, BIAS2 and dBAIS2 also recorded the highest values ​​of Separation Index (more than 1). All indices which calculated from the difference between the before and after fire images gave higher values ​​of the SI, as calculating the difference reduces errors in the classification of burned areas. The BIAS2 index takes advantage of the spectral bands in the Red Edge range provided by Sentinel 2A data and provides an important tool for accurately mapping fire in the Cedar-Fir protected area.

Keywords: Burned Area Indices.; Cedar- Fir Protected Area; Forest Fires; Sentinel 2A; Separation Index.

Full paper in English: pdf