Projection of terrestrial drought evolution and its eco-hydrological effects in China
Yuanhang Yang,Jiabo Yin,Shenglian Guo,Lei Gu,Shaokun He,Jun Wang
DOI: https://doi.org/10.1360/TB-2022-0566
2023-01-01
Chinese Science Bulletin
Abstract:Global warming has changed the water-energy budget and biogeochemical cycle in the Earth's system, and has posed significant impacts on meteorological and hydrological variables such as precipitation, atmospheric humidity and terrestrial water storage (TWS), as well as ecosystem productivity. As drought is one of the most damaging weather-related hazards that challenging the social development and ecosystem heath, it is important to understand its characteristic under climate change. Recently, numerous studies have projected future droughts based on the TWS, but failed to investigate the physical mechanisms behind responses of water-heat fluxes and ecosystem carbon budget to drought events. In this study, the pattern of drought evolution and its eco-hydrological effects in China are systematically investigated from the perspective of physical processes by taking the TWS and atmospheric dynamics into consideration. Based on GRACE/ GRACE-FO and Global Land Data Assimilation System (GLDAS) data from 2002 to 2020, the dry/wet land conditions are identified. Then, the complicated responses of humidity and energy factors (e.g., relative humidity, moisture flux convergence, convective available potential energy, and convective inhibition) to drought events are identified by using observations, ERA5 reanalysis and 20th century reanalysis dataset. Meanwhile, machine learning reconstruction dataset and in situ eddy covariance flux towers data are employed to analyze the anomalies of carbon flux under dry and wet conditions in China. Subsequently, future TWS and drought conditions are projected by using a large set of Global Climate Model (GCM) under three representative concentration pathways (RCP2.6, RCP6.0, and RCP8.5), and impacts of future droughts on ecosystem carbon budget are also explored. The drought conditions are quantified by frequency, duration and severity, which are identified by using the run theory. Finally, the analysis of variance method is used to quantify the uncertainty components sourced from GCM, RCP and terrestrial hydrological models (i.e., global hydrological model, land surface model, and dynamic vegetation model). To characterize the carbon budget process in the ecosystem, we use data of three main carbon flux from both observations and model simulations, i.e., gross primary productivity (GPP), total ecosystem respiration (TER) and net ecosystem productivity (NEP). Results show that the land-atmosphere coupling and drought evolution have complex reciprocal feeding effects. Under the three RCPs, the drought conditions in most areas of China are projected to become more severe, and the frequency, duration and severity of extreme drought events will increase significantly. The duration of extreme droughts in eastern northwest China may double, and the intensity of drought will increase by more than 80% in most areas of China by the end of this century. The GPP, TER and NEP in China will be highly affected by drought events, and the drought regulation of ecosystem carbon sink will be weakened. Higher carbon emissions will lead to higher impacts of droughts on vegetation carbon assimilation, and the decrease of GPP caused by drought is the main reason for the change of terrestrial carbon sink. The NEP in drought events over most landmasses in China is projected to show negative anomalies, especially in future climates. Under future warming climates, the regulation function of ecosystem carbon sink might be weakened by droughts, and the higher carbon emission scenario is often accompanied by more severe impacts of droughts on terrestrial carbon sink. In addition, the total variance of drought indicators and carbon budget variables is mainly sourced from internal variability and model uncertainty. The relative contribution rate of internal variation decreases with the contribution rate of the model, and the scenario uncertainty increases slowly. Our results suggest that more intensified drought magnitudes might challenge the ecosystem sustainable development, highlighting an urgent need to improve the ecosystem resilience to adapt to future warming climates.