Monitoring of vegetation productivity in areas of higher susceptibility to severe fires in mainland Portugal

• Autor(a) principal: Barreirinha, André Miguel Labrincha
• Data de Publicação: 2022
• Tipo de documento: Dissertação
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: http://hdl.handle.net/10773/38599
• topic: GPP; NPP; Vegetation recovery; Droughts; Wildfires

Extreme events such as wildfires and droughts deeply affect the global carbon cycle, since they cause massive degradation of the vegetation greenness. A global and continuous monitoring of vegetation dynamics is essential to analyse the impacts of such events in the vegetation and to understand how it recovers after strong disruptions. Remote Sensing provides information with a wide coverage, high spatial and temporal resolution, being therefore a valuable source of data for these studies. Using this type of data, we present here a study of the vegetation evolution in the Iberian Peninsula during the last 21 years, and give special attention to some large wildfires that occurred in Portugal. First, a climatological study of the Iberian Peninsula was done using MODIS MOD17A3 NPP data. The analysis of yearly and inter-annual mean NPP maps makes clear the areas with more vegetation productivity. A clustering analysis, using the k-means method on the inter-annual NPP variability, allowed the discrimination of 5 regions with distinct vegetation characteristics. The NPP integrated over 3 clusters, located in the northeastern and eastern parts of the Iberian Peninsula, showed a positive trend over the last 20 years. This seems to be associated with changes of the area of each land cover present in the clusters, as revealed by the analysis of the data in the MODIS MCD12Q1 land cover Type 2 product. The integrated NPP time series over Portugal’s mainland and over 4 regions (North, Central, Alentejo and Algarve regions) as well as the evolution of the land covers in the same areas were also analysed. Moreover, Portugal showed a decrease in Evergreen Needleleaf Forests and an increase in Grasslands were observed. A steep decrease in the integrated NPP over Portugal and associated with Evergreen Needleleaf Forest and an accentuated increase of the integrated NPP associated with Grassland were observed around 2017. These changes seemed to be coherent with the severe wildfire season of 2017. However, when the analysis was performed in each region, similar results were obtained in the North, Central, Alentejo and Algarve regions. This suggests more factors should be included to explain the identified changes. To estimate recovery times post-fire, a model that describe exponential decaying anomalies, which was previously applied to NDVI data, was here applied to the more time resolved GPP data from the MODIS MOD17A2 product. The results obtained for the burnt area of Monchique in 2003 are coherent with the values previously obtained when the model was applied to NDVI data. The application of the model to the burnt area of the more recent wildfires of 2017 estimates a recovery time of 39 months for the whole burnt area by the wildfires in October of 2017. When the recovery time of vegetation productivity was estimated as a function of the vegetation type that was present before the wildfire, the areas previously occupied by Savannas show a faster recovery (37 months) of vegetation productivity than the areas that were previously occupied by Evergreen Needleleaf Forests (45 months). Therefore, if the recovery time is estimated for different subareas of the global burnt area their values may differ because the land cover combinations for each subarea may be different. This was verified by applying the model to two subareas of the burnt area in the centre of Portugal in October 2017.