The Gravity Recovery and Climate Experiment (GRACE) and GRACE Follow-On (GRACE-FO) satellite missions have provided a revolutionary tool for measuring changes of the Earth’s gravity field from space. Since GRACE was launched in 2002 (and GRACE-FO in 2018), GRACE/GRACE-FO satellite gravity measurements have been used to study large-scale mass variations in the Earth system, including terrestrial water storage change, groundwater depletion, polar ice sheets and mountain glaciers melting, and the global sea level change. However, validation of GRACE/GRACE-FO satellite gravity observations is challenging because of the lack of other independent observations with comparable spatial scales and temporal samplings that are suitable for validating GRACE/GRACE-FO observations.
Jianli Chen, Department of Land Surveying and Geo-Informatics, The Hong Kong Polytechnic University, Hong Kong
The Caspian Sea is the largest enclosed inland body of water on Earth, with a surface area of ~ 371,000 km2 and located within an endorheic (with no outflow) basin between Europe and Asia. The Caspian Sea level has undergone substantial fluctuations during the past several hundred years, including changes of several meters within a few decades in the recent past. Since around 2005, the Caspian Sea level has experienced a dramatic decline averaging about 10 cm/year. The magnitude is almost 30 times larger than the global mean sea level rise rate of ~ 3 mm/yr during the same period. The amplitude of seasonal Caspian Sea level change also reaches to about 20 cm. Considering the enormously large magnitudes of the Caspian Sea level change combined with the huge area of the Caspian Sea, accurately measured Caspian Sea level changes from satellite altimetry provide a unique well-defined mass change signal that can be used to validate GRACE/GRACE-FO satellite gravity measurements.
Despite the many successful applications of GRACE/GRACE-FO satellite gravity measurements in studying large-scale mass variations in the Earth climate system over the past two decades, one particular challenge for using GRACE/GRACE-FO is the coarse spatial resolution of GRACE/GRACE-FO measurements, which is estimated to be at least a few hundred km. The high-degree and -order gravity coefficients of GRACE/GRACE-FO measurements are dominated by noise and spatial filtering and smoothing are needed in order to extract meaningful mass change signals from GRACE/GRACE-FO measurements. The coarse spatial resolution and needed spatial filtering and smoothing make it difficult to accurately quantify regional mass change using GRACE/GRACE-FO measurements. To overcome this challenge, some post-processing methods, such as Scale Factor (SF), Buffer Zone (BF), and Forward Modeling (FM) have been developed to help minimize or reduce the leakage error in GRACE/GRACE-FO mass change estimates.
Figure 1 shows comparisons between monthly Caspian Sea level changes observed by satellite altimeter and two separate estimates from GRACE/GRACE-FO satellite gravimetry based on the spherical harmonic (GSM) solutions and mascon (MC) solutions provided by the Center for Space Research (CSR), University of Texas at Austin. Leakage corrections for the GRACE/GRACE-FO GSM and MC solutions are implemented using the SF and BF methods, respectively. Terrestrial water and steric effects on the two GRACE/GRACE-FO estimates are also considered using land surface model predictions and World Ocean Atlas data (See Chen et al., 2024 for details).
With appropriate treatments of spatial leakage error and terrestrial water and steric effects, both GRACE/GRACE-FO spherical harmonic and mascon solutions derived Caspian Sea level changes agree remarkably well with independent satellite altimetry observations over a broad range of time scales. The estimated average linear trend from satellite altimetry is -7.55±0.17 cm/yr, while the two GRACE/GRACE-FO estimated trends are -7.50±0.18 and -7.70±0.20 cm/yr, respectively. The average annual amplitude from satellite altimetry is 17.75±1.28 cm, compared with GRACE/GRACE-FO estimates of 18.44±1.34 and 17.05±1.49 cm. The annual phases of satellite altimetry and GRACE/GRACE-FO also agree remarkably well, especially for the GSM solutions. Very good agreement can be found at interannual time scales as well.
The Caspian Sea level change has been proven to be a nearly perfect candidate for independent validations of GRACE/GRACE-FO satellite gravity measurements. It also offers a unique testing base for evaluating the effectiveness of the different data processing methods and strategies developed for GRACE/GRACE-FO applications. Appropriate leakage corrections play a critical role in reducing errors in GRACE/GRACE-FO estimated regional mass change. In the present case, without leakage correction, GRACE/GRACE-FO can only recover less than half of the Caspian Sea level change. Special considerations are also needed when dealing with GRACE/GRACE-FO observed climate change signals with different spatial characteristics and in different geophysical settings.
Reference
Chen, J.L., Wilson, C.R., Seo, K.W., Cazenave, A., Wang, S.Y., Li, J., Nie, Y. (2024). Validation of GRACE/GRACE-FO Solutions Using Caspian Sea Level Change, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, Vol. 17, 15885-15899, DOI: 10.1109/JSTARS.2024.3448488.