Assessment of high spatial resolution satellite imagery for monitoring riparian vegetation: riverine management in the smallholding

Abstract

Riverine habitats are essential ecotones that bridge aquatic and terrestrial ecosystems, providing multiple ecosystem services. This study analyses the potential use of high-resolution satellite imagery, provided by the WorldView-2 satellite, in order to assess its viability for monitoring riparian ecosystems. It is performed by calculating the riparian strip quality index (RSQI) and calibrating it with the riparian quality index (QBR). The methodology was implemented in the Umia River, which is characterised by elevated anthropogenic pressures (located in the northwest of Spain). The results obtained by the method have a 92% of veracity and a kappa coefficient of 0.88. The average quality value obtained for the RSQI index was 71.57, while the average value for the QBR was 55.88. This difference could be attributed to the fact that the former does not differ between autochthonous and non-autochthonous vegetation. The areas with more accurate mapping corresponded to stretches of vegetation with optimal cover (80–50%), with good connectivity with the adjacent forest ecosystem and few or no presence of invasive plants. The worst-scoring sites had the next characteristics: low connectivity (< 10%), low forest cover (< 10%) and a higher presence of invasive plants. The degradation of vegetation could be explained by the presence of agriculture and deficient land use rationing caused by the type of ownership of the study area. The application of this index through satellite images will facilitate the environmental governance of multiple ecosystems and in special riparian ecosystems, obtaining a quick and objective methodology, easily replicable in other basins. Graphical Abstract

The Use of Sentinel-2A Images to Estimate Potential Flood Risk With A Multi-Index Approach in The Mempawah Watershed

“…The changes in the results of class users' accuracy reached 90.909%, while those that did not change were 83.333%. A satellite evaluation rate of > 70% was described as good (Gallego et al, 2014;Rivas-Fandiño et al, 2023). This means that when the accuracy of the interpretation results and change detection approach was good, the producer's accuracy in each class was also good with a value > 88.333%.…”