This paper presents a novel hybrid method named HW-EZ for short-to medium-term forecasting of ΔLOD and UT1-UTC up to 90 days ahead. This method relies on the LOD time series with zonal tidal signals removed, the adaptive Holt-Winters (HW) additive algorithm, and the effective angular momentum (EAM) functions derived from global atmospheric, oceanic, and land water models. The dominant ΔLOD variations are estimated using EAM and tides, while the remaining ones are modeled by HW-EZ with an 8-year sliding window. Numerical validation shows that HW-EZ reduces UT1-UTC mean absolute error (MAE) by 13.77%–67.55% relative to IERS Bulletin A over 1–30-day forecasts. Using 6-day EAM forecasts maintains nearly equivalent accuracy. Comparisons with the Second Earth Orientation Parameters Prediction Comparison Campaign (2nd EOP PCC) confirm that HW-EZ outperforms the average of more than 50 international methods and is competitive with state-of-the-art schemes. The HW-EZ method achieves robust ultra-short-term performance with UT1-UTC MAE ≤0.158 ms for 1–6-day forecasts, making it suitable for real-time geodetic and navigation applications.

Under climate change, conventional droughts that accumulate over long periods and flash droughts that intensify within weeks increasingly co-occur, yet existing drought indices typically focus on a single aspect of the hydrological system and cannot simultaneously resolve long-term cumulative deficits and short-term rapid intensification. To address this limitation, we adopt a meteorological–hydrological composite drought index (GP-MSDI), which integrates terrestrial water storage (TWS) inverted from Global Navigation Satellite System (GNSS) observations with precipitation data through a Copula joint-probability model, and couple it with a three-dimensional spatiotemporal connectivity algorithm to establish a unified framework for identifying both conventional and flash drought events at monthly and weekly scales. Applying this framework to the western United States from 2005 to 2024, we find that GP-MSDI agrees closely with the U.S. Drought Monitor (USDM) in both temporal evolution and spatial propagation. A total of 12 conventional and 64 flash drought events are identified, with the peak spatial extent of two megadroughts almost covering the entire study area. Approximately 39% of the flash droughts evolved into prolonged drought events. The frequency and persistence of flash droughts are strongly modulated by the long-term hydrological background. The three megadrought episodes each coincided with phases of simultaneously negative Niño 3.4 and Pacific Decadal Oscillation (PDO) indices. The proposed framework is transferable to other regions equippedwith dense GNSS networks where integrated drought products comparable to USDM are not available, providing methodological support for regional water-resource management and drought-risk assessment.

While numerous studies have been dedicated to the determination of Chandler wobble parameters, namely its period TCW and quality factor QCW, most of their results lack self-consistency. We proposed a rigorous approach to achieve self-consistent estimates of TCW and QCW using only geodetic excitations (χobs) derived from observed polar motion (PM) data in 2023 (denoted as CCRLL). However, a study published in 2025 (denoted as YF) criticized CCRLL's approach and argued that either TCW or QCW cannot be optimized without access to geophysical excitations (χgeo), relying on their misleading numerical simulations using real-valued PM transfer functions. Our present study has established theoretical relations between (TCW, QCW), degree-2 Love numbers and complex-valued transfer functions while proving that traditional Earth's rotation theories, employing real-valued transfer functions, are incapable of providing self-consistent estimates of TCW and QCW. This study also demonstrates that either χobs or χgeo can be used to invert for TCW and QCW, but χgeo must be recalculated using the complex transfer functions derived by this study and χobs-based estimates are more reliable. Thus, CCRLL's approach and results are validated while YF's arguments are refuted. Further, some updates and corrections to CCRLL's Love numbers and transfer functions are also presented.

The gravity recovery and climate experiment (GRACE) and its successor GRACE follow-on (GRACE-FO) provide critical insights into global terrestrial water storage anomalies (TWSAs) by precisely measuring changes in intersatellite distances. However, various uncertainty sources cause geographically correlated noise in the GRACE/GRACE-FO Level-2 data, which is manifested as north-south striping noise in TWSA grids and complicates hydrological investigations. To address these challenges, we introduce a novel deep learning framework, the Bayesian optimization convolutional neural network with spatial attention mechanism (BO-CNN-SAM), which directly mitigates systematic errors and recovers TWSA from Level-2 data. The results are as follows: 1) The BO-CNN-SAM model architecture demonstrates superior predictive accuracy and robustness, with the root mean squared error (RMSE) reducing by 142%, 120%, and 93% during training, and 136%, 89%, and 74% during validation, compared to other deep learning structures such as BO-CNN, CNN-SAM, and CNN, respectively. 2) In terms of recovered TWSA, BO-CNN-SAM outperforms other CNN-based models, achieving the lowest RMSE (4.25, 4.40, and 5.50 cm) and the highest Nash-Sutcliffe Efficiency coefficients of 0.78, 0.64, and 0.60 during the training, validation, and testing phase, respectively. 3) The BO-CNN-SAM framework is successfully used to estimate ice sheet mass balance in Greenland and to accurately quantify prolonged drought in the Yangtze River Basin during 2022. Looking ahead, BO-CNN-SAM has the potential to directly recover TWSA from raw GRACE/GRACE-FO Level-1 data to gridded Level-3 TWSA products. This capability opens new opportunities for producing TWSA fields with low latency and enhancing our understanding of hydrological processes.

Accurately recovering the time-variable gravity from GRACE (Gravity Recovery and Climate Experiment) and GRACE-FO (GRACE Follow-On) is crucial for studying global land-ocean water exchange, monitoring regional water budget balance, and predicting local droughts and floods. However, the commonly used method of scale factor for recovering the true time-variable gravity signals has several limitations: different hydrological models yield different scale factors, the "striping" error and high-frequency signals are not considered when applying the same filtering process to hydrological models, and the subjective selection of filter strength. To address these issues, we propose an improved scale factor algorithm that includes: (1) integration multiple hydrological models into a unified set of hydrological data using the Bayesian Three-Cornered Hat method; (2) consideration of the "striping" error and high-frequency signals in the filtering process hydrological models; (3) introduction of evaluation metrics and iteration to determine DDK4 as the optimal filtering solution for GRACE/GRACE-FO level-2 time-variable gravity field data within the DDK1-DDK7 filtering range. The improved scale factor method shows good agreement with the results from CSR Mascon and JPL Mascon, with correlation coefficients of 0.984 and 0.970, and root mean square errors of 2.13 mm and 3.54 mm, respectively. Moreover, the spatiotemporal distribution of the recovered terrestrial water storage anomaly using the improved scale factor closely matches that of CSR Mascon and JPL Mascon in terms of trend, annual amplitude, and phase values.

We develop a novel composite drought index, the GNSS and Precipitation based Multivariate Standardized Drought Index (GP-MSDI), integrating Global Navigation Satellite System (GNSS) and precipitation data to improve drought characterization in Brazil. GP-MSDI combines meteorological and hydrological variables using the Copula function, enabling multidimensional evaluation of drought events. Applied from January 2010 to December 2022, GP-MSDI identified 7-9 drought events across four major basins in Brazil, with Sao Francisco River Basin (SRB) experienced the longest drought duration (116 months, 74.4% of the study period). Compared to other drought indices, GP-MSDI effectively captures compound drought events. Spatial analysis highlighted that the Amazon River Basin (ARB) and eastern Brazil faced more severe droughts. Following the spatial analysis, we introduced the Drought Severity Exposure Risk Index (DSERI), which combines GP-MSDI drought severity with population density to assess socio-environmental vulnerability. DSERI revealed high-risk areas in southeastern and northeastern Brazil, where severe droughts overlap with densely populated regions, underscoring the urgent need for targeted mitigation strategies. This study introduces GP-MSDI as a comprehensive tool for drought monitoring, offering an integrated approach by combining meteorological and hydrological data through geodetic methods, thereby enhancing our understanding of drought dynamics in Brazil.

GNSS-LEO occultation detection is widely used in radio-sounding, climate research, and space exploration because of its global coverage and high vertical resolution. The traditional occultation event prediction method has the problem of high computational cost, while this study reduces the computation time and improves the accuracy by adaptively adjusting the time step. Experimental results show that the adaptive time-step algorithm significantly improves the efficiency of predicting occultation events compared with the Brute-force solution method in the application of different satellite constellations.

Satellite gravimetry (Gravity Recovery and Climate Experiment (GRACE)/GRACE Follow‐On (GFO)) has solved the challenge of monitoring global and basin‐scale terrestrial water storage anomalies (TWSA) at monthly intervals and spatial scales of ∼330 km. However, this spatial resolution limits its ability to capture small‐scale basin water variations, and GRACE/GFO cannot independently distinguish different vertical water components, which further restricts its application in hydrological studies. To address these challenges, we propose a joint inversion downscaling method that combines hydrological simulations and GRACE/GFO observations. Our approach effectively inherits the high spatial resolution patterns from hydrological models and preserves the large‐scale accuracy from the GRACE/GFO measurements. It also allows for the flexible inclusion of mascon groups, reducing dependence on hydrological models and improving the performance in glacier regions, where traditional downscaling methods struggle. We reconstructed high‐resolution (i.e., 0.5°) TWSA and its components in China, for example, groundwater and glacier, revealing their finer spatial trends and localized water mass changes. From 2002 to 2022, our study details TWSA and flux changes in different China's basins, demonstrating marked regional disparities. Groundwater depletion rate in the North China Plain is 2.43 ± 0.18 cm/yr, equivalent to 3.35 ± 0.25 Gt/yr, and glacier mass loss rate in the High Mountain Asia (HMA) is 34.10 ± 1.55 Gt/yr. We also map the spatiotemporal patterns of glaciers in HMA and surface water changes in the endorheic Tibetan Plateau. This work has successfully extracted high‐resolution signals for different vertical water components in China, providing valuable insights for regional water resource management and disaster mitigation.

While the geodetic excitation χ(t) of polar motion p(t) is essential to improve our understanding of global mass redistributions and relative motions with respect to the terrestrial frame, the widely adopted method to derive χ(t) from p(t) has biases in both amplitude and phase responses. This study has developed a new simple but more accurate method based on the combination of the frequency- and time-domain Liouville's equation (FTLE). The FTLE method has been validated not only with 6-h sampled synthetic excitation series but also with daily- and 6-h sampled polar motion measurements as well as χ(t) produced by the interactive webpage tool of the International Earth Rotation and Reference Systems Service (IERS). Numerical comparisons demonstrate that χ(t) derived from the FTLE method has superior performances in both the time and frequency domains with respect to that obtained from the widely adopted method or the IERS webpage tool, provided that the input p(t) series has a length around or more than 25 years, which presents no practical limitations since the necessary polar motion data are readily available. The FTLE code is provided in the form of MatLab function.

The Gravity Recovery and Climate Experiment (GRACE) and GRACE Follow-On (GRACE-FO) missions have enabled consistent production of monthly gravity field solutions by international institutes, contributing to the International Centre for Global Earth Models. Each institute employs distinct processing strategies, yielding varied estimates of terrestrial water storage (TWS). In this study, we employ statistical collocation techniques (Total assessment ratio, TAR) to assess and compare the performance of GRACE TWS data products (2003.03 similar to 2014.03) and GRACE-FO TWS data (2018.06 similar to 2022.11). For GRACE TWS, the TAR values are as follows: COST-G (0.15), ITSG (0.83), APM-SYSU (0.85), CSR (0.91), JPL (0.93), GFZ (0.94), Tongji (0.96), HUST (1.08), SUST (1.18), CNES (1.37), and AIUB (1.41). Similarly, for GRACE-FO TWS, the TAR values are COST-G (0.15), JPL (0.81), ITSG (0.96), CSR (0.97), GFZ (1.06), and CNES (1.41). Furthermore, our comparison across basin sizes and climatic regions reveals that COST-G exhibits lower uncertainty and larger signal-to-noise ratios in TWS, making it particularly noteworthy for its utility. Conversely, other single solutions that depict long-term trends and annual amplitudes demonstrate comparable values across various basin sizes, climatic regions, and specific areas.

The Antarctic Ice Sheet (AIS) has been losing ice mass and contributing to global sea level rise (GSLR). Given its mass that is enough to cause ∼58 m of GSLR, accurate estimation of mass balance trend is critical for AIS mass loss monitoring and sea level rise forecasting. Here, we present an improved approach to reconciled solutions of mass balance in AIS and its regions from multiple contributing solutions using the input-out, altimetric, and gravimetric methods. In comparison to previous methods, such as IMBIE 2018, this approach utilizes an adaptive data aggregation window to handle the heterogeneity of the contributing solutions, including the number of solutions, temporal distributions, uncertainties, and estimation techniques. We improved the regression-based method by using a two-step procedure that establishes ensembled solutions within each method (input-output, altimetry, or gravimetry) and then estimates the method-independent reconciled solutions. For the first time, 16 contributing solutions from 8 Chinese institutions are used to estimate the reconciled mass balance of AIS and its regions from 1996 to 2021. Our results show that AIS has lost a total ice mass of ∼3213±253 Gt during the period, an equivalent of ∼8.9±0.7 mm of GSLR. There is a sustained mass loss acceleration since 2006, from 88.1±3.6 Gt yr−1 during 1996–2005 to 130.7±8.4 Gt yr−1 during 2006–2013 and further to 157.0±9.0 Gt yr−1 during 2014–2021. The mass loss signal in the West Antarctica and Antarctic Peninsula is dominant and clearly presented in the reconciled estimation and contributing solutions, regardless of estimation methods used and fluctuation of surface mass balance. Uncertainty and challenges remain in mass balance estimation in East Antarctica. This reconciled estimation approach can be extended and applied for improved mass balance estimation in the Greenland Ice Sheet and mountain glacier regions.

For time-variable satellite gravity solutions of GRACE and GRACE-FO in terms of spherical harmonics coefficients, the Scale Factor (SF) is often used to recover the close to “true” signal of Terrestrial Water Storage (TWS) anomalies. However, the conventional SF method has some limitations that may hinder its effectiveness, including: (1) their dependency on input hydrological models that may lead to divergent estimations of SFs; (2) the arbitrary choice of filter strength, which may not be representative in different regions; (3) limited consideration of SF in the temporal dynamics (the conventional SF was fixed value) for monthly varied TWS. Here, we propose a new Dynamic SF method to overcome these limitations and increase accuracy of the restored global TWS changes. This method involves: (1) the Bayesian Three-Cornered Hat (BTCH) method is applied to merge three sets of hydrological products into an optimal hydrological dataset to be used for estimating a unique SF, (2) the anisotropic DDK3 filter (found to be numerically optimal) is applied to suppress the correlated noise, and (3) an iterative Kalman filter process is formulated and implemented to estimate monthly Dynamic SF corresponding to monthly global TWS fields. The Dynamic SF outperformed an ordinary SF method in terms of the Root Mean Squared Error (RMSE), the Mean Absolute Error (MAE), and the Signal to Noise Ratio (SNR), which were found to be improved by 30.8%, 32.2%, and 41.3%, respectively. Moreover, the recovered TWS using the Dynamic SF method showed a good agreement with the GRACE/GRACE-FO RL06 mascon solutions of the CSR and that of JPL regarding the long-term trend, seasonal, and interannual variations.

On September 5, 2022, the Luding M6.8 earthquake occurred in the Moxi-Shimian segment of the Xianshuihe fault, coinciding with the historical ruptured zone of the 1786 Moxi earthquake. Its seismogenic environment provides a foundation for comprehending the mechanism of the earthquake and its future hazard. In the Moxi-Shimian segment, we establish a series of near-field Global Navigation Satellite Systems (GNSS) stations to enhance the spatial resolution of observational data for the inversion of the interseismic kinematic parameters. In this study, with an elastic screw dislocation model constrained by GNSS observations, the slip rate of the Moxi-Shimian segment is estimated to be 10.9 ± 1.0 mm/yr, while the locking depth is 15.7 ± 6.2 km. Additionally, we utilize a block-dislocation model to invert the interseismic fault coupling along the Kangding-Moxi-Shimian segment. The result indicates a gradual deepening of the locking depth along the section from Kangding to Shimian. The coseismic rupture of the 2022 event occurred within the high coupling regions in the Kangding-Moxi-Shimian segment, which indicates that the rupture kinematics in this event might be controlled by the interseismic deformation. The seismic moment accumulated within the ruptured zone of the Luding earthquake since 1786 ranges in [1.42–3.40] × 1019 N·m, which is significantly greater than the seismic moment released during the 2022 event. As a result, we infer that the Luding earthquake released only a portion of the accumulated energy within the original rupture zone since 1786, indicating that the 2022 event has not caused a complete rupture in the Moxi-Shimian segment. Consequently, there remains a substantial seismic hazard in this area.

The period T CW and quality factor Q CW of the Chandler wobble (CW) as well as polar motion (PM) transfer functions are all determined by the Earth’s layered structure, mass distribution, elasticity, rheology and energy dissipation, via the Earth’s dynamic figure parameters and complex degree-2 Love numbers. However, most previous studies used geophysical excitations derived from real-valued PM transfer functions to invert for T CW and Q CW , thus leading to results that are not self-consistent. By separating the observed PM into the freely decaying CW and the excited PM, a traverse-based method is proposed to search values of T CW and Q CW that can fit both sides simultaneously, yielding the self-consistent estimates of T CW = 430.4 mean solar days and Q CW = 130. This implies the degree-2 tidal Love number k = 0.35011 − 0.00226i and load Love number k' = − 0.36090 + 0.00233i, and the PM transfer functions T NL = 1.80001 − 0.00692i (non-loading) and T L = 1.15040 − 0.00023i (loading) valid at the Chandler period.

Terrestrial water storage (TWS) is a pivotal component of the global water cycle, profoundly impacting water resource management, hazard monitoring, and agriculture production. The Gravity Recovery and Climate Experiment (GRACE) and its successor, the GRACE Follow-On (GFO), have furnished comprehensive monthly TWS data since April 2002. However, there are 35 months of missing data over the entire GRACE/GFO observational period. To address this gap, we developed an operational approach utilizing singular spectrum analysis and principal component analysis (SSA-PCA) to fill these missing data over mainland China. The algorithm was demonstrated with good performance in the Southwestern River Basin (SWB, correlation coefficient, CC: 0.71, RMSE: 6.27 cm), Yangtze River Basin (YTB, CC: 0.67, RMSE: 3.52 cm), and Songhua River Basin (SRB, CC: 0.66, RMSE: 7.63 cm). Leveraging two decades of continuous time-variable gravity data, we investigated the spatiotemporal variations in TWS across ten major Chinese basins. According to the results of GRACE/GFO, mainland China experienced an average annual TWS decline of 0.32 ± 0.06 cm, with the groundwater storage (GWS) decreasing by 0.54 ± 0.10 cm/yr. The most significant GWS depletion occurred in the Haihe River Basin (HRB) at −2.07 ± 0.10 cm/yr, significantly substantial (~1 cm/yr) depletions occurred in the Yellow River Basin (YRB), SRB, Huaihe River Basin (HHB), Liao-Luan River Basin (LRB), and Southwest River Basin (SWB), and moderate losses were recorded in the Northwest Basin (NWB, −0.34 ± 0.03 cm/yr) and Southeast River Basin (SEB, −0.24 ± 0.10 cm/yr). Furthermore, we identified that interannual TWS variations in ten basins of China were primarily driven by soil moisture water storage (SMS) anomalies, exhibiting consistently and relatively high correlations (CC > 0.60) and low root-mean-square errors (RMSE < 5 cm). Lastly, through the integration of GRACE/GFO and Global Land Data Assimilation System (GLDAS) data, we unraveled the contrasting water storage patterns between northern and southern China. Southern China experienced drought conditions, while northern China faced flooding during the 2020–2023 La Niña event, with the inverse pattern observed during the 2014–2016 El Niño event. This study fills in the missing data and quantifies water storage variations within mainland China, contributing to a deeper insight into climate change and its consequences on water resource management.

Contemporary research on Greenland surface mass balance (SMB) is largely focused on the characteristics of decadal or longer trends, periodic oscillations, and acceleration. However, the specific components of the SMB such as snowfall (SF), rainfall (RF) and runoff (RU), and their corresponding temporal and spatial variability remain poorly understood. Here, we explore the respective contributions of SF, RF, and RU to the seasonal and transient crustal deformations of Greenland during the past two decades using GPS network and satellite gravimetry (GRACE) datasets, and regional climate model output. Our study unraveled that the largest annual vertical displacement caused by precipitations is in southeastern Greenland, reaching 7.27 mm. The largest surface displacement caused by RU is in western Greenland, reaching 19.82 mm. Ice mass gain/loss in Greenland shows a clear correlation between latitude and temperature, with greater variations in the south compared to the north. The transient deformation signals in Greenland mainly manifested in terms of abrupt subsidence in 2010, followed by uplift in 2014. The 2014 uplift can mainly be attributable to the combined effect of SF, RF, and RU. The largest transient signal occurs in the southeast subregions, with peak-to-peak amplitude exceeding 10 mm. Transient crustal deformation is mainly caused by precipitation in southeastern Greenland, while the contribution of RU dominates most of the time and in most subregions. We find that even though RF is increasing due to an increasingly warmer climate, its effect on SMB is still negligible, when compared with SF and RU. In some subregions and some periods, SF could become the primary contributor to transient SMB variations in Greenland.

By taking into account the variable free polar motion (PM) known as the Chandler wobble (CW) and irregular forced PM excited by quasi-periodic changes in atmosphere, oceans and land water (described by the data of effective angular momenta EAM), we propose a short-term PM forecast method based on the Holt-Winters (HW) additive algorithm (termed as the HW-VCW method, with VCW denoting variable CW). In this method, the variable CW period is determined by minimizing the differences between PM observations and EAM-derived PM for every 8-year sliding timespan. Compared to the X- and Y-pole forecast errors (ΔPMX and ΔPMY) of the International Earth Rotation and Reference Systems Service (IERS) Bulletin A, our results derived from operational EAM can reduce ΔPMX by up to 38.4% and ΔPMY by up to 34.3% for forecasts ranging from 1 to 30 days. Further, we prove that using EAM forecast instead of operational EAM in the HW-VCW method can achieve similar accuracies.

The Intergovernmental Panel on Climate Change's (IPCC) Special Report on the Ocean and Cryosphere in a Changing Climate (SROCC) concluded that the Antarctica and Greenland ice sheets were major contributors to global sea-level rise. Satellite gravimetry, e. g. the Gravity Recovery and Climate Experiment (GRACE) and the GRACE Follow-on (GRACE-FO) missions, provides the direct observations of large-scale ice mass loss. However, there is a one-year data gap spanning from July 2017 to May 2018 between GRACE and GRACE-FO. Fortunately, operating from November 2013, the Swarm mission with a constellation of three LEO (Low Earth Orbit) satellites can be used to recover global large-scale temporal gravity fields. Meanwhile, the ARIMA-MC (Autoregressive Integrated Moving Average Model-Monte Carlo) method also makes forecasting the missing time-varied gravity possible. The main objective of this study is to determine ice mass variations in Antarctica and Greenland during the one-year gap between GRACE and GRACE-FO. Our results indicated that Swarm's results were comparable to the forecast from the ARIMA-MC method in space-time. From April 2002 to March 2020, the rates of ice mass loss were -119 +/- 23 Gt.a(-1) and -259 +/- 20 Gt.a(-1) over Antarctica and Greenland, respectively, which was equivalent to the global sea level rise at the rate of similar to 0. 33 mm.a(-1) and similar to 0. 72 mm.a(-1), respectively. The ice loss rate of Wilkes Land in Antarctica during 2010-2020 was accelerated by 10 times comparing to the period of 2002-2009. The major abrupt ablation of Greenland in the summer of 2019 was connected to the NAO (North Atlantic Oscillation) event.

Past redistributions of the Earth’s mass resulting from the Earth’s viscoelastic response to the cycle of deglaciation and glaciation reflect the process known as glacial isostatic adjustment (GIA). GPS data are effective at constraining GIA velocities, provided that these data are accurate, have adequate spatial coverage, and account for competing geophysical processes, including the elastic loading of ice/snow ablation/accumulation. GPS solutions are significantly affected by common mode errors (CMEs) and the choice of optimal noise model, and they are contaminated by other geophysical signals due primarily to the Earth’s elastic response. Here, independent component analysis is used to remove the CMEs, and the Akaike information criterion is used to determine the optimal noise model for 79 GPS stations in Antarctica, primarily distributed across West Antarctica and the Antarctic Peninsula. Next, a high-resolution surface mass variation model is used to correct for elastic deformation. Finally, we use the improved GPS solution to assess the accuracy of seven contemporary GIA forward models in Antarctica. The results show that the maximal GPS crustal displacement velocity deviations reach 4.0 mm yr−1, and the mean variation is 0.4 mm yr−1 after removing CMEs and implementing the noise analysis. All GIA model-predicted velocities are found to systematically underestimate the GPS-observed velocities in the Amundsen Sea Embayment. Additionally, the GPS vertical velocities on the North Antarctic Peninsula are larger than those on the South Antarctic Peninsula, and most of the forward models underestimate the GIA impact on the Antarctic Peninsula.

Accurately monitoring spatio-temporal changes in lake water levels is important for studying the impacts of climate change on freshwater resources, and for predicting natural hazards. In this study, we applied multi-mission radar satellite altimetry data from the Laurentian Great Lakes, North America to optimally reconstruct multi-decadal lake-wide spatio-temporal changes of water level. We used the results to study physical processes such as teleconnections of El Niño and southern oscillation (ENSO) episodes over approximately the past three-and-a-half decades (1985–2018). First, we assessed three reconstruction methods, namely the standard empirical orthogonal function (EOF), complex EOF (CEOF), and complex independent component analysis (CICA), to model the lake-wide changes of water level. The performance of these techniques was evaluated using in-situ gauge data, after correcting the Glacial Isostatic Adjustment (GIA) process using a contemporary GIA forward model. While altimeter-measured water level was much less affected by GIA, the averaged gauge-measured water level was found to have increased up to 14 cm over the three decades. Our results indicate that the CICA-reconstructed 35-year lake level was more accurate than the other two techniques. The correlation coefficients between the CICA reconstruction and the in situ water-level data were 0.96, 0.99, 0.97, 0.97, and 0.95, for Lake Superior, Lake Michigan, Lake Huron, Lake Erie, and Lake Ontario, respectively; ~7% higher than the original altimetry data. The root mean squares of errors (RMSE) were 6.07 cm, 4.89 cm, 9.27 cm, 7.71 cm, and 9.88 cm, respectively, for each of the lakes, and ~44% less than differencing with the original altimetry data. Furthermore, the CICA results indicated that the water-level changes in the Great Lakes were significantly correlated with ENSO, with correlation coefficients of 0.5–0.8. The lake levels were ~25 cm higher (~30 cm lower) than normal during EI Niño (La Niña) events.