GRACE-based investigation of large-scale land-atmosphere interactions

Author: Ajiao Chen

Chen, Ajiao, 2022 GRACE-based investigation of large-scale land-atmosphere interactions, Flinders University, College of Science and Engineering

Terms of Use: This electronic version is (or will be) made publicly available by Flinders University in accordance with its open access policy for student theses. Copyright in this thesis remains with the author. You may use this material for uses permitted under the Copyright Act 1968. If you are the owner of any included third party copyright material and/or you believe that any material has been made available without permission of the copyright owner please contact with the details.


Land-atmosphere interactions encompass complex surface processes that exchange energy and matter between land and the atmosphere, which play important roles in modulating variations in climate. Prediction on future climate change calls for more precise prediction models. Improving the representation of physical processes of land-atmosphere interactions and the availability of key variables for characterizing those processes could help reduce uncertainties in the prediction models, and consequently make contribution to extreme weather forecasting and natural disasters prevention.

Terrestrial water storage (TWS, includes surface water, soil moisture, groundwater, snow, and ice) constitutes a significant memory component within the climate system. However, in Australia, the driest inhabitant continent, there is still a lack of investigation on the long-term TWS variation pattern. In addition, soil moisture as the most variable component of TWS has strong interactions with vegetation and near-surface temperature, but investigations on those interactions have been impeded by the scarcity of soil moisture observations.

The long-term wetting/drying pattern in Australia was investigated in this thesis by applying the Gravity Recovery and Climate Experiment (GRACE) satellite derived TWS anomaly and extended datasets. A seesaw pattern of TWS variation between eastern and western Australia was revealed: eastern Australia gaining water, while western Australia is losing water, and vice versa. This phenomenon is resulted from a combination of effects from large-scale climate mode and dynamic vegetation and soil moisture interactions. It highlighted the bidirectional effects between surface vegetation and land water conditions, but such knowledge of Australia remained poorly understood. Results of this thesis for the first time indicated that non-linear interactions between vegetation and TWS occurred in 58% of the area of Australia. Those new findings partly improved our understanding of physical processes in Australia’s land-atmosphere interactions.

On the other hand, this thesis proposes the first use of wavelet decomposed GRACE TWS as a proxy of soil moisture to investigate its relationship with air temperature anomaly/hot extremes at the global scale. Compared to raw TWS, decomposed TWS showed improved skill in explaining temperature variability. It is because that the decomposed components could reflect different roles of moisture at different soil depths in the soil moisture-temperature coupling. The wavelet decomposed TWS also performed better than other commonly used soil moisture proxies (i.e., precipitation relevant index, products derived from land surface model and microwave remote sensing technology). Besides, by using the decomposed TWS to represent local moisture deficit, it played a more important role in influencing hot extreme occurrences in regions with a total area 1.6 times as large as the area strongly influenced by global temperature change during the study period 1985–2015. The results suggested that local land management is essential for combating hot extreme expansion in regions with strong land-atmosphere coupling, and global measures for reducing emissions are required in the face of increasing greenhouse gas forcing.

In summary, this thesis improved the knowledge of land-atmosphere interactions at continental and global scales through further investigation on TWS variation pattern and its relationships with vegetation and temperature. This thesis also suggested a useful soil moisture proxy, i.e., the wavelet decomposed GRACE TWS, that can be applied to examine other processes in land-atmosphere interactions and to evaluate the performance of land surface models.

Keywords: land-atmosphere interaction, GRACE, terrestrial water storage, seesaw, soil moisture-temperature coupling, hot extremes

Subject: Earth Sciences thesis

Thesis type: Doctor of Philosophy
Completed: 2022
School: College of Science and Engineering
Supervisor: Huade Guan