Terrestrial Carbon Cycle and its Feedback at the Regional Scale

Abstract

Mapping ecosystem carbon across different scales and comparing estimates from various systems is essential, both in its own right and for understanding the increasing exchange of atmospheric carbon dioxide (CO2) between the atmosphere and biosphere. While the global terrestrial carbon sink helps mitigate the accumulation of atmospheric CO2, this process is contingent upon climate and ecosystem factors. The fundamental biophysical mechanisms governing ecosystem-carbon-climate interactions and their feedback mechanisms remain highly uncertain. Despite extensive efforts to track changes in ecosystem dynamics and attribute them to environmental factors using sophisticated data platforms, there is intense debate about whether studying regional carbon budgets can reconcile discrepancies in carbon flux calculations and improve global carbon budget estimates. Significant carbon variability is associated with large uncertainties stemming from Land Use Changes (LUC), resulting in a regional carbon source at seasonal to interannual scales, although without long-term positive or negative feedback. In the face of rapid LUC, continuous monitoring of carbon variability is crucial to understand India's role as a carbon sink in the global budget. Unlike other global regions, limited observational networks in India have hindered efforts to capture dynamics and fluctuations in the Indian carbon budget. However, recurrent Earth observation systems have facilitated monitoring carbon flux variability from diurnal to decadal timescales and from local to global spatial scales, aiding research in understanding ecosystem traits and subsequent carbon variability. Integrating predictive Earth System Models (ESMs) with diverse data streams reveals the sensitivity of carbon fluxes to various global environmental drivers across diverse climate and vegetation gradients. Focusing on understanding India's regional carbon dynamics in recent history, this dissertation employs in-situ, remote sensing, and process-based models to emphasise the interaction of regional carbon dynamics with multiple drivers. Flux variability, in both magnitude and pattern, differs across ecosystems but demonstrates a strong consistency among datasets. Integrating eddy covariance observations with remote sensing data emphasises the importance of synergistic use of multivariate datasets in understanding ecosystem productivity across temporal scales. India's diverse flora results in varying carbon uptake across biomes, with tropical ecosystems serving as dominant carbon storage hubs. However, the regional carbon cycle is reshaped by multiple environmental drivers, subsequently influencing climate patterns. Considering atmospheric aerosols as a hindrance, a remote sensing process-based model, the Carnegie Ames Stanford Approach (CASA), was employed to examine the potential effect of aerosol load on ecosystem productivity across diverse agroclimatic zones of India. Carbon flux sensitivity varies across ecosystems, with pronounced positive and negative feedback effects observed over forest and cropland ecosystems. To explore the complex dynamics of India's carbon uptake under various forcing scenarios over the past century, the Community Earth System Model (CESM) was utilised. This model traces how rising atmospheric CO2 concentrations and climate changes influence India's net land sink. Principal climate drivers were considered to identify their roles in potentially triggering long-term shifts in Indian ecosystem functionality as a carbon sink. Although the analysis indicates India's historical role as a carbon sink, asymmetries in decadal and seasonal trends from the integrated model ensemble suggest the potential for future terrestrial carbon loss, particularly in forest-based ecosystems. Explicit analysis by this research advances our understanding of India's carbon sink dynamics and underscores the sensitivity of carbon uptake to various environmental challenges. It emphasises the urgent need for adaptive strategies in the face of these challenges. Considering the ecosystem-specific sensitivities between carbon uptake and environmental drivers, future efforts incorporating divergent data platforms into process-based models within specific gradients will significantly enhance our understanding and prediction of future carbon uptake at regional and global scales.