| dc.description.abstract | Terrestrial ecosystems play an important role in the global carbon cycle by naturally taking up carbon from the atmosphere through photosynthesis and storing this carbon as biomass and soil carbon. There exists a strong, yet complex and uncertain relationship between the terrestrial ecosystems and the climate system. Inadequate knowledge on the regional terrestrial carbon cycle dynamics translates into large uncertainties in the estimation of global terrestrial carbon stocks and fluxes and in turn the climate projections as well. It is important to gain a better understanding of the terrestrial carbon dynamics at a regional level, in order to reduce the uncertainty in the terrestrial carbon estimates at the global scale (LeQuéré et al., 2017, Canadell et al., 2011). The Indian region is very important but relatively unexplored in terms of terrestrial carbon studies. However, significant environmental changes have occurred in the 20th century that has affected the terrestrial carbon dynamics of the Indian region considerably. The present research provides a model based assessment of the trends and variability in terrestrial carbon stocks and fluxes over India. We have used a multi-model based approach to understand the regional dynamics of the carbon cycle in the historical period and to predict the future terrestrial carbon dynamics over India.
In the first part of the thesis (Chapter 3), we assess the trends and variability in the land carbon uptake in India during the period 1901-2010 using outputs from nine land surface models that are a part of the TRENDY model inter-comparison project. Outputs from two simulations were used - the S1 simulation where only atmospheric CO2 concentration is varying and the S2 simulation where both the climate and atmospheric CO2 are varying. The simulation S2 represents approximately the actual historical evolution of the system and the S1 simulation is used to make comparisons with S2 to obtain the individual effect of climate change (S2-S1). The changes in S1 simulation with time is used to infer the effect of increasing CO2. Our analysis in this chapter is focused on the trends and variability in Net Primary Productivity (NPP), Net Ecosystem Productivity (NEP) and the Net Ecosystem Exchange (NEE) over India in the historical period. The TRENDY multi-model mean NPP shows a positive trend of 2.03% per decade during the historical period from 1901-2010 over India. The multi-model based estimate of the cumulative NEE is 0.613 ± 0.1 Pg C during 1901-2010, indicating that the Indian terrestrial ecosystem was neither a strong source nor a significant sink during this historical period. Inter-annual variation in the terrestrial carbon variables are strongly driven by local precipitation changes and remote drivers such as ENSO and Indian Ocean Dipole (IOD) do not have a strong influence on the terrestrial carbon variables over India. The water use efficiency over the region is assessed along with the terrestrial carbon variables, and it shows an increase of about 25% over the 110-year period. However, since our conclusion is based on a set of offline land models further investigations are necessary for assessing the robustness of our estimates based on the TRENDY models. Thus, in the second part of our study we assess the terrestrial carbon dynamics over the Indian region, using a single Dynamic Global Vegetation Model (DGVM) driven by climate data from multiple climate models.
In the second part of the thesis, we use a Dynamic Global Vegetation Model (DGVM) forced with climatic inputs from 24 CMIP5 climate models, to assess the trends and variability in terrestrial carbon stocks and fluxes over India for the historical period and for two Representative Concentration Pathway (RCP) scenarios - the RCP 4.5 scenario and the RCP 8.5 scenario. Outputs from 24 Coupled Model Intercomparison Project phase 5 (CMIP5) climate models are used to force the DGVM for more statistically significant results. The DGVM used is the Lund-Potsdam-Jena DGVM (LPJ-DGVM) and is run on a grid cell basis at 0.5° latitude and longitude resolution. For the historical period, we find that the range in the terrestrial carbon fluxes obtained from LPJ-DGVM forced with 24 CMIP5 climate models are comparable with the fluxes obtained from the TRENDY multiple land surface model based assessments. The ensemble mean NPP from the 24 LPJ-DGVM simulations increases by11.3% in the historical period. In the RCP 4.5 scenario, NPP increases by 20% and in the RCP 8.5 scenario, by 62% through the end of the 21st century compared to the 1996-2005 baseline. The vegetation carbon and soil carbon show an increasing trend in the historical period as well as the RCP 4.5 and the RCP 8.5 scenario. The ensemble mean NEE from the 24 LPJ-DGVM simulations over India, has a positive mean value through the historical period, the RCP 4.5 scenario and the RCP 8.5 scenario. The cumulative NEE for the Indian region over the historical period is positive and concur with the TRENDY multi-model results from the previous chapter although differing in the magnitude of source. The cumulative NEE is positive for the RCP 4.5 scenario and is slightly negative in the RCP 8.5 scenario, indicating the Indian land mass becomes a source of carbon to the atmosphere in the RCP 4.5 scenario and sink for the RCP 8.5 scenario.
In the third part of the thesis, we assess the role two key drivers, climate change and CO2, for the trends and variability in terrestrial carbon variables over India in the historical period and in the two RCP scenarios. We find that the terrestrial ecosystems over India are a source of carbon when the climate is varying with warming increasing through the historical periods and reaching maximum in the RCP 8.5 scenario. Thus, the climate effect alone makes the Indian land mass a source of carbon to the atmosphere in the historical period as well as the RCP 4.5 scenario and the RCP 8.5 scenario. The contribution of the increasing concentrations of CO2 in the atmosphere is positive on the terrestrial carbon stocks and fluxes and we find that the productivity and hence the terrestrial carbon stocks increase as a result of the CO2 fertilization effect in the historical period, the RCP 4.5 and RCP 8.5 scenarios. The cumulative changes in NEE as a result of the climate-effect, the CO2-effect and the net effect are assessed and we find that in the historical period from 1900-2005, climate-effect is a stronger driver than the CO2-effect, driving the Indian region to be a source for carbon as a net effect. In the RCP 4.5 scenario, from 2006-2100, the climate-effect continues to be a stronger driver than the CO2-effect, driving the Indian region to be a net source for carbon. In the RCP 8.5 scenario, however, the CO2 fertilization effect is a stronger driver and the terrestrial ecosystems are a small net sink for carbon.
To summarize, this thesis makes a multi-model based assessment of the long term trends and variability in terrestrial carbon stocks and fluxes over the Indian region for the 20th century and predicts the future changes during 21st century. Understanding the past, current and future changes in terrestrial ecosystem dynamics helps plan better policy interventions to preserve the terrestrial ecosystems from further degradation and in turn mitigate the effects of climate change in the Indian region. | en_US |
