Please use this identifier to cite or link to this item:
|Title:||Evaluation of Potential Erosion Risk after Forest Fires Applying Spectral Mixture Analysis Incorporating Spectroradiometric Field Data|
|Authors:||O'BRIEN VICTORIA; BOETTCHER KRISTIN; MEHL WOLFGANG; PALUMBO Ilaria; BARBOSA FERREIRA PAULO|
|Citation:||Proccedings of the 5th International Conference on Forest Fire Research|
|Type:||Articles in periodicals and books|
|Abstract:||Land degradation due to deforestation after forest fires is increasingly recognized as a problem. The impact on the ecosystem is very difficult to quantify as the collection of field measurements is highly time consuming and it is usually limited to a restricted area. This study tested the ability of spectral mixture analysis applied to LandsatTM5 imagery to produce realistic and meaningful endmember fractions for the study of changes in land surface conditions after forest fires. The imagery was used to evaluate the difference between the pre- and post fire dataset in order to determine the extent of change from burning/combustion. SMA gives the opportunity to categorise the scene into various sub-areas and is guiding the endmember selection for a better adjustment of the model to different surface types. An iterative feedback process can automatically select appropriate endmembers for each pixel. The unmixing process can be supported by implemented data from other information sources and improved at each stage. The analysis was performed on a large burned area (3500 ha) in the commune of Genova, Italy, where a fire occurred on the 14-18th of February 2005. Moreover, spectral measurements were taken with the ASD Fieldspec II to develop a spectral library as an accurate reference source of spectral reflectance, which was integrated in the unmixing model. Furthermore, field data on physical parameters such as fire severity, vegetation cover, rock cover and soil cover were collected in the study site for further validation of the spectral unmixing. The overall purpose of the study was to extract the highest amount of variability that is related to our interest, changes in the presence/absence of vegetation cover and the related erosion risk. The outcome of the comparison of the classified pre- and post-fire scenes can give early stage assessment on the risk of erosion due to the degree of change. The preliminary results obtained in this study show promising perspectives on the use of SMA for the analysis of vegetation status and erosion risk after forest fires; however, in order to fully validate the results obtained in this study, it will be necessary to continue this analysis on longer time frame.|
|JRC Institute:||Sustainable Resources|
Files in This Item:
There are no files associated with this item.
Items in repository are protected by copyright, with all rights reserved, unless otherwise indicated.