Dokument

Summary:
Global forest cover has suffered a dramatic reduction during recent decades, especially in tropical regions, which is mainly due to human activities caused by enhanced population pressures. Nevertheless, forest ecosystems, especially tropical forests, play an important role in the carbon cycle functioning as carbon stocks and sinks, which is why conservation strategies are of utmost importance respective to ongoing global warming. In South America the highest deforestation rates are observed in Ecuador, but an operational surveillance system for continuous forest monitoring, along with the determination of deforestation rates and the estimation of actual carbon socks is still missing. Therefore, the present investigation provides a functional tool based on remote sensing data to monitor forest stands at local, regional and national scales. To evaluate forest cover and deforestation rates at country level satellite data was used, whereas LiDAR data was utilized to accurately estimate the Above Ground Biomass (AGB; carbon stocks) at catchment level. Furthermore, to provide a cost-effective tool for continuous forest monitoring of the most vulnerable parts, an Unmanned Aerial Vehicle (UAV) was deployed and equipped with various sensors (RBG and multispectral camera). The results showed that in Ecuador total forest cover was reduced by about 24% during the last three decades. Moreover, deforestation rates have increased with the beginning of the new century, especially in the Andean Highland and the Amazon Basin, due to enhanced population pressures and the government supported oil and mining industries, besides illegal timber extractions. The AGB stock estimations at catchment level indicated that most of the carbon is stored in natural ecosystems (forest and páramo; AGB ~98%), whereas areas affected by anthropogenic land use changes (mostly pastureland) lost nearly all their storage capacities (AGB ~2%). Furthermore, the LiDAR data permitted the detection of the forest structure, and therefore the identification of the most vulnerable parts. To monitor these areas, it could be shown that UAVs are useful, particularly when equipped with an RGB camera (AGB correlation: R² > 0.9), because multispectral images suffer saturation of the spectral bands over dense natural forest stands, which results in high overestimations. In summary, the developed operational surveillance systems respective to forest cover at different spatial scales can be implemented in Ecuador to promote conservation/ restoration strategies and to reduce the high deforestation rates. This may also mitigate future greenhouse gas emissions and guarantee functional ecosystem services for local and regional populations.

Das Dokument ist im Internet frei zugänglich - Hinweise zu den Nutzungsrechten