Spatio-temporal variability of vegetation phenology and its drivers at global scale

Abstract

Phenology is the study of the timing of periodic life-cycle events in plants and animals, as well as how seasonal changes in climatic factors affect these events. In recent decades, temperature has increased, leading to changes in the timing of leaf phenophases and the lengthening of the photosynthetically growing season. These changes in vegetation phenology have an impact on the water and carbon cycle; the lengthening of the growing season has increased the carbon uptake of terrestrial ecosystems offsetting part of the atmospheric carbon from human emissions. However, future responses of vegetation phenology to warming are uncertain, raising concerns about the ability of vegetation to offset atmospheric carbon emissions. The main objective of the thesis was to characterize the spatial and temporal variability of land surface phenology and link this variability to climate drivers in a context of global warming. The research focused on understanding how rising temperatures might change the climate factors affecting vegetation phenology. The first two chapters deal with the methodology for estimating land surface phenology. In Chapter 1, we propose a new method for estimating land surface phenology in cloud-based platforms that can be applied to raw time series without the need for time series preprocessing. In Chapter 2, we present 10-meter resolution maps for the continental scale, emphasizing the importance of spatial resolution. Chapters 3 to 5 cover the impact of three climate factors, temperature, light, and water availability, on the vegetation phenology. In Chapter 3, results suggest that frozen soils constraint vegetation activity, and vegetation resume photosynthesis closely after soil thawing. Chapter 4 points at the climate constraints on carbon uptake phenology, particularly the limitation of radiation in temperate and cold regions in the Northern Hemisphere. These findings suggest that the start of the growing season may still advance, although at a slower pace, while radiation restrains photosynthesis in autumn, preventing the further delay of the end of season with future warming. Chapter 5 studies another factor, the occurrence of heatwaves and droughts, that advances the end of the growing season. In this chapter, we show evidence of early leaf shedding using 10-meter resolution satellite data, and link this phenomenon to high temperature and aridity conditions. These findings reveal that early leaf shedding is more recurrent and widespread than previously reported. The last chapter, Chapter 6, is a compendium of techniques and knowledge gained in previous chapters. We model the land surface phenology estimated with the technique describe in Chapter 1 and using a model based on the climate factors studied in Chapters 3-5. We analyze the climate constraints on vegetation phenology at the global scale using a novel technique that explains the relationship between sun-induced fluorescence and climate factors in a machine learning model. The findings of the thesis demonstrate different spatial constraints of temperature, light, and water availability at the beginning and end of the growing season, suggesting that vegetation phenology will respond differently to future climatic warming depending on the location and type of vegetation. In some regions, rising temperatures may not translate into a lengthening of the growing season because of radiation and water constraints.

Date of Award	9 Sept 2022
Original language	English
Supervisor	Josep Peñuelas Reixach (Director) & Ten Aleixandre Verger (Director)