Remote Sensing of Forest Fire at Mediterranean Level: the Challenge of Transferring Information across scales
We present and discuss results using remotely-sensed imagery for studying forest fires in the Mediterranean region. We focus on two aspects: processing of high-resolution imagery for detailed information pre- and post-fire vegetation cover, and the information provided by coarse resolution sensors. The first aspect is important because forest management on one hand, and the advances on fire ecology on the other, require data of increasing higher quality. The second aspect is critical to integrate different sources of imagery taking advantage of the temporal information contributed by coarse resolution sensors, and to correct estimates of burnt area derived from coarse imagery for regional studies. We successfully applied image segmentation and model-based clustering to the high resolution imagery to detect the fire scar and map pre-fire conditions. We also studied the spectral responses of the different covers to post-fire conditions. Coarse resolution imagery was used to show discrepancies as high as 70% on burnt area estimates and to study the temporal dynamics of burnt versus not-burnt vegetation.
Bibliographic Reference: Paper presented: 3rd International Conference on Forest Fire Research, Coimbra (PT), November 16-21st, 1998
Availability: Available from DG XIII as Paper EN 41906 ORA
Record Number: 199910325 / Last updated on: 1999-03-12
Original language: en
Available languages: en