Remotely sensed monitoring of land surface albedo and ecosystem dynamics

Abstract

The earth is under unprecedented pressure, which is reflected in rapid ecosystem changes around the globe. Over just the past three decades, the earth has lost over 178 million hectares of its forests. The rapidly growing evidence of the loss of resilience in ecosystems (i.e., recovery from disturbances slows down) due to climate change has become a global concern. Our limited knowledge of ecosystem dynamics and their key parameters, such as albedo, has also hindered our ability to manage ecosystems appropriately. The main aim of this dissertation is to contribute to elucidating ecosystem dynamics by exploiting remotely sensed satellite data. Additionally, this dissertation aims to explore the dynamics of albedo (reflectivity) in response to forest structure (forest density, tree cover, and leaf area index) variations and forest disturbances (fire and drought). To address specific research questions discussed throughout this dissertation, various study sites extending from local to global scales are considered. The results showed that using an appropriate ecosystem state variable that can represent the state of an ecosystem makes it is possible to achieve a reliable and timely evaluation of wetlands or forests state change. To that end, a new remotely sensed index called the modified vegetation water ratio (MVWR) was developed which improved the ability to understand the dynamics of wetlands. A new approach was also developed based on incorporating local spatial autocorrelation. The efficiency of this approach in measuring the state of drought-affected forests was demonstrated. To provide a convenient implementation of the presented approaches, a new R package, "stew", was developed. Investigating the dynamics of forest albedo on disturbances revealed that precipitation and burn severity only weakly explained the temporal dynamics of albedo. In contrast, it was shown that the number of fire events and the leaf area index strongly explained temporal albedo dynamics. According to the results, although fires could lead to abrupt decreases in the temporal dynamics of albedo, droughts caused abrupt increases in the temporal dynamics of albedo. Finally, a global-scale study was conducted to explore the links between forest structure and albedo during the peak growing season. The results demonstrated that forest structure might significantly explain albedo in most of the forests around the world. It was also found that the response of the shortwave albedo (300–5000 nm) to the variations of the leaf area index was always positive. The first map representing the links between forest structure and albedo was provided which highlights the importance of the role of forests in modulating albedo on a global scale. It is expected that the implications of the results of this dissertation contribute to future climate mitigation plans, the monitoring of ecosystem dynamics, and sustainable development in general.