Remote Sensing in Forest Fires: The Case of Bandipur

On 23 February 2019, news articles reported fire spreading in the Bandipur Tiger Reserve across the Himavad Gopalaswamy Betta range in the south Indian state of Karnataka. The reserve is bordered by farms in the north and the Mudumalai and Wayanad wildlife sanctuaries in the south. Bandipur, Mudumalai, and Wayanad fall under the critical tiger habitat spread across Karnataka, Tamil Nadu, and Kerala that is estimated to host one-eighth of the tiger population worldwide.

Figure 1: Bandipur Tiger Reserve: 21/02/2019 vs. 26/02/2019

Bandipur tiger reserve is a dry deciduous forest with a history of forest fires. We wanted to get a closer look at the forest fire using remote sensing. To get a clear picture, we acquired pre-fire (11/02/2019) and post-fire (26/02/2019) open source Sentinel-2 satellite imagery. We created a true colour composite (TCC) to view the spread of fire as shown in Figure 2. We used indices on the TCC that could give us additional insights. In particular, the Ashburn Vegetation Index (AVI) is used in the growth stage to measure green vegetation. AVI measures the height of the plant/tree from the tillering stage to the harvest stage. It is also referred to as the Agricultural Vegetation Index for this reason. AVI can also be used to measure the height of trees from a fully grown stage to the stage after burning or clearing.

Figure 2: True Colour Composite of the Bandipur Forest Area on February 26, 2019

We applied AVI to the satellite images, which showed us the cleared or burnt agricultural pockets. Figure 3 shows clearing/burning in Bandipur area with values increasing from light to dark orange (not to be confused with the TCC). Dark orange areas are active/severe fire affected areas that give information on the spread, direction, and magnitude of the fire. Light orange refers to areas that have been burned but are not active or severe in comparison to dark orange areas.

Figure 3: Ashburn Vegetation Index (AVI) applied on the Bandipur Area

When we zoom in around Bandipur forest area in Figure 4, we see that agricultural plots adjacent to the forest (Shivapura, Siddapura & Melukamanahalli) have similar AVI value (Ashburn VI) to that of the burnt area across Bandipur. This implies that farmers set their plots ablaze to prepare fields for the next growing season in spite of awareness drives against it. In villages around Gopalswamy Betta, awareness drives were conducted to educate farmers to avoid setting their fields on fire as they touch the border of the Bandipur Forest. Figure 5 shows Ashburn VI overlaid with Google maps as the base layer. The fire in Bandipur spread southwards to cause damage in the Mudhumalai & Wayanad wildlife sanctuaries, part of the critical tiger habitat.

Figure 4: Zoomed in satellite image of the area around Bandipur
Figure 5: Ashburn VI over Google basemap

Remote sensing and geospatial analytics play a pivotal role in tackling an ongoing fire event. This provides information on the magnitude, spread, and direction of fire that can help during disaster management. During the planning stage of fire management practices, however, implementation of remote sensing intelligence can play an even greater role as it can equip administrative bodies to intelligently execute strategies to prevent a fire event and mitigate its catastrophic effects.

Puneeth Shankar | Lead Remote Sensing Scientist, GeoSpoc

Leave a Reply

Your email address will not be published. Required fields are marked *