Wed, Aug 04, 2021:On Demand
Background/Question/Methods: The loss of biodiversity is one of the most important consequences of the current global warming scenario. A first step for its conservation is to identify climatic microrefugia: thermally stable areas where certain organisms can withstand climatic changes. In this project, we have used thermal sensors attached in an unmanned aerial vehicle (UAV; Anafi Thermal recording in RGB and thermal infrared at centimetric resolutions) for the identification of microrefugia across six environmentally heterogeneous areas of the Pyrenees, a southern European mountain range. The UAV-thermal data has allowed us to estimate land surface temperatures (LST) at two contrasted moments of the day and two seasons in 2020. For each centimetric pixel we have estimated the length of the thermal range it experienced in each season (an index of thermal stability), and we have created a seasonal layer including the ones within the highest 10% values. The overlap of both layers has allowed us to identify the most stable pixels for each site. UAV-LST data have been validated with a network of miniaturized thermal sensors (iButton®) in the same areas. Furthermore, using photogrammetric techniques from the UAV-RGB data, we have generated digital surface models (DSM) to derive geo-environmental variables (slope and aspect), and we have digitalized the main land covers (grassland, screes, rocky, and forested areas) from the RGB orthophotos.
Results/Conclusions: Measurements of temperatures from UAV and miniaturized sensors correlated rather well (R2: 0.7741). The index of thermal stability of each pixel was modelled using Generalized Linear Models as a function of geo-environmental variables and the land cover, all of them having a significant effect on it. The results have shown that the most stable sites are forest patches regardless of the aspect, and north-faced cliffs. On the contrary, the less stable areas have been grasslands in south-faced slopes, rocky surfaces and few screes sectors. This is the first study to date that has used a UAV in complex areas to produce thermal landscapes and identify climatic microrefugia. The potential of this method is highly promising not only because of the extraordinary fine resolution UAVs provide, but also because it allows to work in remote or inaccessible areas. Our study demonstrated the value of combining LST and information from DSM for the precise identification of climatic microrefugia at organism scale in topographically complex areas.
Results/Conclusions: Measurements of temperatures from UAV and miniaturized sensors correlated rather well (R2: 0.7741). The index of thermal stability of each pixel was modelled using Generalized Linear Models as a function of geo-environmental variables and the land cover, all of them having a significant effect on it. The results have shown that the most stable sites are forest patches regardless of the aspect, and north-faced cliffs. On the contrary, the less stable areas have been grasslands in south-faced slopes, rocky surfaces and few screes sectors. This is the first study to date that has used a UAV in complex areas to produce thermal landscapes and identify climatic microrefugia. The potential of this method is highly promising not only because of the extraordinary fine resolution UAVs provide, but also because it allows to work in remote or inaccessible areas. Our study demonstrated the value of combining LST and information from DSM for the precise identification of climatic microrefugia at organism scale in topographically complex areas.