| In recent years,global warming has increased the frequency and intensity of wildfires,which has put tremendous pressure on forest ecosystems.Restoring forests has been identified as the most effective solution to extreme fire and climate change.Studying the response relationship between early regeneration of post-fire vegetation communities and climate environmental factors is of great scientific and practical significance for enhancing the overall resilience of forests.This article takes the"Bobcat Wildfire"in the San Gabriel Mountains of California,USA as an example,and studies the effect of wildfire disturbance and post-fire climate change on the early regeneration of forest and shrub vegetation communities based on 13 Sentinel-2 phase time-series images(4×13 52 images in total).The main conclusions and innovative research achievements of this article are as follows:(1)This study conducted research on the extraction of burned areas and the mapping of burn severity using K-Means clustering algorithm combined with differential normalized vegetation index(d NDVI)and differential normalized burn ratio(d NBR),and the d NBR threshold of the European Forest Fire Information System(EFFIS).The results indicate that the classification method combining the K-Means clustering algorithm with the d NBR index yields the highest overall accuracy(99.1%)for extracting burned area,which is approximately 4.5%higher than the accuracy obtained by the EFFIS empirical threshold,this method can replace the empirical threshold in this study;Moreover,this method exhibits a high level of consistency with EFFIS in terms of the classification results for areas with high levels of burn damage and above.Analyzing the temporal recovery dynamics of burn severity,found that the area would need at least 6.91 years to recover to an undamaged/low damage level.(2)An innovative vegetation community image secondary optimization classification method was employed,which combined a space-spectrum cooperative K-Means clustering algorithm with a support vector machine(SVM)classification algorithm.The accuracy of the method was validated by producing a confusion matrix with 389 vegetation sample points.The results indicate that the overall accuracy(OA)of the vegetation community classification was approximately 92.3%to 99.5%,and the Kappa coefficient was 0.88 to 0.98,after the pre-classification processing with the K-Means clustering algorithm,secondary optimization classification with the SVM algorithm and the optimal index(NBR)combination.The accuracy was improved by approximately 6.6%compared to the accuracy before optimization,achieving a further improvement of the image classification accuracy.(3)A temporal monitoring and analysis was conducted on the dynamic changes of early vegetation coverage and regeneration area in the woody and shrubby vegetation community.The results indicated that the accumulated burned area of vegetation by the Wildcat wildfire was approximately 220.94 km~2.Among them,the total forest canopy area of the forest community decreased by 81.4%(about 65.49 km~2),while the shrub community decreased by 52.5%(about 155.01 km~2).After 367 days of natural growth,the total regenerated area of the forest community was approximately 7.85 km~2,with an average regeneration rate of 0.66 km~2 per month;the total regenerated area of the shrub community was approximately 58.29 km~2,with an average regeneration rate of4.86 km~2 per month.It would take approximately 10 years for the overall vegetation coverage to recover to the pre-fire level,with about 8 to 9 years needed for the forest community and about 3 years or even less needed for the shrub community.(4)A quantitative analysis was conducted to examine the mechanisms relationship between climate,topographic factors,burn severity and early regeneration of forest and shrub vegetation communities after a wildfire.The results show that the three climate factors significantly affect the regeneration of both forest and shrub vegetation communities(R~2:0.42-0.88).Among them,sunshine duration and rainfall are the most important factors for the regeneration of forest community,while the shrub community requires a significant amount of rainfall.The regeneration trend of forest vegetation on sunny slopes and shrub vegetation on shaded slopes is obvious and exhibits strong fluctuations,while forest vegetation on shaded slopes and shrub vegetation on sunny slopes exhibit a gradient towards stable regeneration.The optimal slope for vegetation regeneration is between 15-35 degrees.The severity of burning has altered the original competitive relationship and regeneration rate of the forest and shrub vegetation communities.Tolerable low-burn injuries can promote the regeneration of forest vegetation but inhibit that of shrub vegetation.Forest vegetation has the ability to resist some fire disturbance,while shrub vegetation is more adapted to high-intensity fire environments due to its resilience and resistance.This study provides a reference for supporting the implementation of regional forest management strategies and enhancing the resilience of post-fire vegetation communities in a targeted manner. |
PDF Full Text Request