Understanding Carbon Uptake Using Multi-Scale Novel Remote Sensing Techniques

Abstract

Drylands cover 40% of the global terrestrial area and support roughly 2 billion people for grazing and cropping. With hotter, drier climate forecasts for dryland, in particular, there are new challenges in understanding dryland responses and vulnerability. This has emerged as a priority research frontier because dryland carbon dynamics have been identified as disproportionately important in regulating interannual variability of atmospheric CO2 concentrations. Remote sensing can be used to monitor multiple aspects of vegetation dynamics as well as its sensitivity to climate to improve dryland carbon uptake estimations. The thesis integrated existing cutting-edge remote sensing proxies to reduce uncertainties in dryland carbon uptake estimations from global to ecosystem scale. The results suggested that at a global scale, solar-induced fluorescence (SIF) captures gross primary productivity (GPP) phenology more accurately than the normalized difference vegetation index (NDVI) and vegetation optical depth (VOD). In dryland regions, integrating near-infrared reflectance of terrestrial vegetation (NIRv), SIF and heterogeneity will improve satellite-based GPP estimates. In a semi-arid grassland ecosystem, integrating SIF and the photochemical reflectance index (PRI) will improve the understanding of photosynthesis during extreme droughts. Overall, this research integrated multiple available remotely sensed proxies to improve understanding of vegetation response to climate change across scales, across heterogeneous regions, and during extreme droughts. My research has the potential to revolutionize the way we monitor dryland vegetation productivity through integrating different and independent proxies or adjusting GPP models’ inputs. These understandings will also provide a valuable vegetation productivity map to inform adaptive ecosystem management strategies in support of climate mitigation in drylands.