Antarctic ice cover quality abnormal event identification method and system

Abstract

The invention provides an Antarctic ice cover quality abnormal event identification method and system, and belongs to the technical field of polar region environment monitoring, and the method comprises the steps: obtaining the GRACE / FO gravity data of each month in a set time period; obtaining ice cover surface material balance data, and dividing an accumulation period and an ablation period of the South Pole ice cover AIS; error evaluation is carried out based on GRACE / FO gravity data, a significance threshold value is set, and months with significant quality change are obtained through screening; and based on the division result of the accumulation period and the ablation period, obtaining the antarctic ice cover quality abnormity event from the months with significant quality change. Based on a daily GRACE / FO gravitational field solution, instant ice cover quality abnormity caused by an extreme climate event can be monitored, and effective tracking of a short-term AIS quality change event is realized.

Description

Technical Field

[0001] This invention belongs to the field of polar environment monitoring technology, specifically relating to a method and system for identifying anomaly events in the quality of the Antarctic ice sheet. Background Technology

[0002] The Antarctic Ice Sheet (AIS) is the largest land ice system on Earth, accounting for approximately 90% of global land ice. It covers an area of ​​about 13.5 million square kilometers, with an average thickness of 2,450 meters and a total volume of nearly 26.5 million km³, storing almost 70% of the Earth's freshwater. The complete melting of the AIS would cause global sea levels to rise by approximately 58 meters, posing a serious risk to global human societies and natural ecosystems. Driven by global warming, the AIS has been melting at an accelerated pace in recent decades, substantially impacting the stability of the global climate system through complex interactions with the atmosphere and oceans.

[0003] Precise monitoring of Antarctic ice sheet (AIS) mass changes is fundamental to elucidating its spatiotemporal variability and the physical mechanisms controlling its mass budget. These efforts provide crucial data support for improving sea-level rise predictions, refining ice sheet dynamics modeling, and advancing our understanding of climate feedback. Enhanced observation and analysis of AIS mass balance also contribute to a more comprehensive understanding of polar environmental change and the interactions between the ice sheet, atmosphere, and ocean systems.

[0004] The existing GRACE / FO mission provides continuous satellite gravity observations, enabling the tracking of AIS mass changes and providing data support for rapid responses to extreme weather events. However, the commonly used monthly GRACE / FO gravity field cannot reliably separate events driven by special weather / hydrological processes from normal seasonal variations or long-term trends. Furthermore, due to the strong seasonal cycles inherent in ice sheets (weight gain during accumulation periods and weight loss during ablation periods), even if mass changes are observed, it is difficult to determine whether they constitute anomalies in a statistical and physical sense. Summary of the Invention

[0005] To address the problem of identifying anomalies in the Antarctic ice sheet mass, this invention provides a method for identifying anomaly events in the Antarctic ice sheet mass.

[0006] To achieve the above objectives, the present invention provides the following technical solution: A method for identifying anomaly events in Antarctic ice sheet mass includes the following steps: Acquire GRACE / FO gravity data within a specified time period; the specified time period is measured on a monthly, 10-day, or daily scale. Obtain surface mass balance (SMB) data of the Antarctic ice sheet, and divide the accumulation and ablation periods of the Antarctic ice sheet AIS based on the SMB data; Monthly mass change error assessment is performed based on the GRACE / FO gravity data. A significance threshold is set based on the monthly mass change error assessment results, and months exceeding the significance threshold are marked as months with significant mass changes. If a month with significant mass changes experiences mass loss during the accumulation period or mass increase during the melting period, it is judged as an Antarctic ice sheet mass anomaly event.

[0007] Preferably, the division of the Antarctic ice sheet AIS into accumulation and ablation periods based on the SMB data includes the following steps: Use the average of the SMB time series data for all months within a specified time period as the ID data; The quality anomaly data is obtained by subtracting the ID data from the SMB time series of all months within a set time period; Based on the aforementioned quality anomaly data, the annual cycle AC of AIS quality is calculated. If the monthly AC is positive, it is the accumulation period; if it is negative, it is the ablation period.

[0008] Preferably, the monthly mass change error assessment based on the GRACE / FO gravity data specifically includes the following steps: Calculate the average monthly GRACE / FO gravity data from different datasets; wherein, the different datasets are monthly gravity monitoring data from different institutions; The variance between different datasets is calculated using the triangular hat method, and the RMS error is calculated based on the variance. The first difference of the monthly average series is obtained to obtain the monthly quality change difference series and the corresponding RMS error for each month.

[0009] Preferably, the significance threshold is specifically: ; in, Monthly quality changes for the AIS system. This represents the RMS error.

[0010] If the relationship between quality change and error satisfies the above formula, and the GRACE / FO observation data for adjacent months are not missing, then a significant increase or decrease in quality is considered to exist in that month; conversely, if the quality change is submerged in 3 times the RMS error, it is not considered a significant change.

[0011] Preferably, the method further includes determining the driving proportion of the quality anomaly event results, specifically including the following steps: The original time series of AIS mass changes is extracted from the GRACE / FO gravity data; the long-term trend, annual period and semi-annual period of the original time series are fitted to obtain the residual series after deducting the trend term and period term from the original mass time series, and the difference value is calculated. The contribution ratio is calculated based on the difference value of the original quality time series and the quality difference value after removing the trend and periodic components. If the contribution ratio is positive, it means that the quality change becomes more obvious after removing the long-term trend and periodic fluctuations, and the quality change is dominated by short-term processes. If the contribution ratio is negative, it means that the trend and periodic components have explanatory power for the quality change of the month, and the quality change is regarded as the result of long-term climate forcing.

[0012] This invention also provides a system for identifying anomaly events in the Antarctic ice sheet, specifically including: The data module is used to acquire GRACE / FO gravity data within a set time period; the set time period is measured on a monthly, 10-day, or daily scale.

[0013] An anomaly identification module is used to acquire the surface mass balance (SMB) data of the Antarctic ice sheet, and to divide the Antarctic ice sheet AIS into accumulation and ablation periods based on the SMB data; to perform monthly mass change error assessment based on the GRACE / FO gravity data, and to set a significance threshold based on the monthly mass change error assessment results, marking months exceeding the significance threshold as months with significant mass changes; if a month with significant mass changes experiences mass loss during the accumulation period or mass increase during the ablation period, it is judged as an Antarctic ice sheet mass anomaly event.

[0014] The present invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps described in the method for identifying anomalies in Antarctic ice sheet quality.

[0015] The present invention also provides a computer-readable storage medium storing a computer program that, when loaded by a processor, can execute the steps described in the method for identifying anomalies in Antarctic ice sheet quality.

[0016] The method for identifying anomaly events in Antarctic ice sheet mass provided by this invention has the following beneficial effects: This invention clearly delineates the ice sheet accumulation and ablation periods based on ice sheet surface mass balance data, establishing a normal seasonal mass change benchmark. This provides a clear physical basis for defining "abnormal events" (loss during accumulation / increase during ablation), avoiding misjudgments based solely on numerical fluctuations. Error assessment is performed using GRACE / FO gravity data, and a significance threshold is set for the error assessment results. Quantifying the error establishes an objective standard for statistical significance of the observed values. Based on the seasonal mass change benchmark and data trends, months with abnormal Antarctic ice sheet mass changes are identified. This diagnoses genuine physical anomalies, rather than simply data fluctuations. Attached Figure Description

[0017] To more clearly illustrate the embodiments and design schemes of the present invention, the accompanying drawings required for this embodiment will be briefly described below. The drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a technical roadmap for identifying anomaly events in the Antarctic ice sheet in an embodiment of the present invention.

[0019] Figure 2 This is a graph showing the AIS quality variation of GRACE / FO data at three different time resolutions in this invention. The fitted curves, trends, annual period amplitudes, and trend charts in the graph are all results from GRACE / FO data at three resolutions during the period from April 2002 to June 2017.

[0020] Figure 3 This is the division of the AIS accumulation period and ablation period in the implementation of this invention. Figure 3 (a) The effect of removing emissions from the original SMB time series; Figure 3 (b) Annual cycle calculation after removing the influence of ice flow.

[0021] Figure 4 The months in which significant changes in AIS quality were identified during the implementation of this invention.

[0022] Figure 5 The months in which the quality of the invention underwent significant changes or abnormalities.

[0023] Figure 6 This is a comparison of three-resolution GRACE / FO data during four abnormal events in an embodiment of the present invention. Figure 6 (a) refers to the quality increase event in January 2012 (Event 1); Figure 6 (b) refers to the quality increase event in October 2004 (Event 2); Figure 6 (c) refers to the quality loss event in July 2004 (Event 3); Figure 6 (d) represents the quality loss event (Event 4) in September 2003. The points in the figure represent the time span covered by the original data, and the curves represent the monthly average series.

[0024] Figure 7 This is a comparison of daily GRACE / FO solutions and daily precipitation during four abnormal events in the implementation of this invention. Figure 7 (a) refers to the quality increase event in January 2012 (Event 1); Figure 7 (b) refers to the quality increase event in October 2004 (Event 2); Figure 7(c) refers to the quality loss event in July 2004 (Event 3); Figure 7 (d) represents the quality loss event (Event 4) in September 2003. The red and green shaded areas in the figure represent positive and negative quality anomalies, respectively. For the daily GRACE / FO solution, the annual period from 2002 to 2012 was extracted and a 5-day moving average was applied.

[0025] Figure 8 This shows the spatial distribution of AIS quality anomalies during abnormal event 3 in the implementation of this invention. Figure 8 (a) represents July 2004 minus the GRACE quality anomaly in June 2004; Figure 8 (b) represents an SMB anomaly (July 2004 minus June 2004). Figure 8 (c) is the subsurface quality anomaly obtained by subtracting SMB from GRACE (i.e., panel a minus panel b).

[0026] Figure 9 This is the GNSS load deformation response to the extreme precipitation event in the AIS region in December 2011, as implemented in this invention. Among them, Figure 9 (a) shows the diurnal spatial distribution of 5-day extreme precipitation and the locations of 16 GNSS stations. Figure 9 (b) shows the time series of vertical displacement of the GNSS station.

[0027] Figure 10 This is the diurnal evolution of atmospheric circulation characteristics during extreme precipitation events in the implementation of this invention. Figure 10 (a) to Figure 10 (e) represents the daily average geopotential height and total precipitation at 500 hPa from December 16 to 20, 2010. Figure 10 (f) to Figure 10 (j) represents the daily average specific humidity at 850 hPa and the total precipitation during the same period; Figure 10 (k) to Figure 10 (o) represents the daily average sea level pressure and total precipitation during the same period.

[0028] Figure 11 This illustrates the spatiotemporal relationship between extreme precipitation and ARs in the implementation of this invention. Figure 11 (a) shows the spatial correspondence between the IVT transport path and extreme precipitation on December 16, 2011; Figure 11 (b) to Figure 11 The figure (u) shows the evolution of the IVT from 00:00 on December 15, 2011 to 09:00 on December 17, 2011. The black dashed boxes in the figure represent the evolution trajectory of ARs, the blue dashed boxes represent the landing locations of ARs, and the red dashed boxes represent the peak times of AR intensity.

[0029] Figure 12 This is a flowchart of a method for identifying anomaly events in the Antarctic ice sheet according to the present invention. Detailed Implementation

[0030] To enable those skilled in the art to better understand and implement the technical solutions of the present invention, the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. The following embodiments are only used to more clearly illustrate the technical solutions of the present invention and should not be construed as limiting the scope of protection of the present invention.

[0031] Example This invention provides a method for identifying anomaly events in the Antarctic ice sheet mass, such as... Figure 1 As shown, the specific steps include: I. Obtaining observational and verification data.

[0032] S11: Observational data refers to GRACE / FO (GravityRecovery and Climate Experiment and Follow-On) data for each month, every 10 days, and each day within a specified time period.

[0033] Monthly GRACE / FO data utilizes three GRACE / FO mascon datasets: a 0.25°×0.25° mascon solution (RL0603) from the University of Texas Center for Space Research (UTCSR); a 0.5°×0.5° solution (RL06.3_v04) from NASA's Jet Propulsion Laboratory (JPL); and a 0.5°×0.5° solution (RL06v2.0) from NASA's Goddard Space Flight Center (GSFC). All datasets have a monthly time resolution, spanning from 2002 to 2024. The solutions in these datasets essentially correspond to the data's settlement results.

[0034] The 10-day and daily GRACE / FO data used the CNES_GRGS_RL05 10-day gravity field solution provided by the National Center for Space Studies (CNES) and the ITSG-grace2018 daily solution developed by the Institute of Geodesy at Graz University of Technology (ITSG). Both datasets are published in the form of spherical harmonic coefficients. The 10-day solution has a maximum degree and order of 90, covering DOY210 from 2002 to DOY244 from 2024, while the daily solution has a maximum degree and order of 40, covering DOY91 from 2002 to DOY181 from 2017.

[0035] The acquired data was preprocessed, and all low-order spherical harmonic coefficients were linearly interpolated to match the time resolution of their respective datasets. Specifically, the 1-degree coefficients were taken from TN-13_GEOC_CSR_RL0603, C20 and C30 from TN-14_C30_C20_GSFC_SLR, and C21, S21, C22, and S22 from UTCSR. The interpolated coefficients were then used to replace the corresponding terms in the original solution. The CNES 10-day solution employed truncated singular value decomposition (TSVD) to regularize higher-order noise components, thereby improving the signal-to-noise ratio and stability. Kalman filtering was applied to the ITSG daily solution to apply dynamic constraints, significantly suppressing high-frequency noise. To maintain the integrity of the original signal as much as possible, the GIA of both datasets was corrected using the ICE6G-D model to ensure consistency with the monthly mascon solution. The resulting gravity field was then converted into Global Equivalent Water Height (EWH) data.

[0036] S12: Validation data includes GNSS time series and auxiliary data.

[0037] Vertical displacement time series from 16 GNSS stations provided by the Nevada Geodetic Laboratory (NGL) were used. These data generally cover the period from 2010 to 2012. Due to data gaps in some records, only relatively complete segments of the record were retained for each station to ensure temporal continuity. Since the analysis involves estimating long-term trends and seasonal variations, each selected time series spans at least two years. Hector software was used to correct offsets and remove outliers from the GNSS time series, and loading products provided by the German Research Centre for Geosciences (GFZ) were used to further remove the effects of non-tidal atmospheric loads (NTAL) and non-tidal ocean loads (NTOL). Furthermore, the thermal expansion effect was corrected using the model of Yanetal (2009). The final processed GNSS time series were used for independent validation in subsequent analyses.

[0038] The supporting data include surface mass balance (SMB) and precipitation data from regional climate models, meteorological reanalysis data, and vertical integrated water vapor transport (IVT).

[0039] Surface Mass Balance (SMB) and Precipitation from Regional Climate Models: To improve the stability of the analysis, three monthly SMB and precipitation datasets generated by regional climate models (MAR and RACMO) were used. MAR version 3.12, developed by the University of Liège, identifies an anomaly in Antarctic ice sheet mass events, covering the period from 1979 to 2021 and performing well in simulating surface mass on the AIS. The RACMO model, developed by the Royal Netherlands Meteorological Institute, includes two versions: RACMO2.3p2 and RACMO2.4p1, covering 1979–2022 and 1979–2023, respectively. Both versions are driven by ERA5 reanalysis as lateral boundary conditions and have undergone continuous improvements in handling microphysical processes and snow-ice interactions. All three datasets contain coupled surface processes and provide a comprehensive representation of surface mass balance changes on the AIS.

[0040] Meteorological Reanalysis: To examine the relationship between AIS quality change events and atmospheric driving factors, two reanalysis datasets, ERA5 and MERRA-2, were used. ERA5, produced by the European Centre for Medium-Range Weather Forecasts (ECMWF), provides a spatial resolution of approximately 31 km and a temporal resolution of hourly values. Variables extracted from ERA5 include total precipitation, 500 hPa geopotential height, 850 hPa specific humidity, mean sea level pressure, and wind fields at 500 hPa and 850 hPa. MERRA-2, developed by NASA's Office of Global Modelling and Assimilation (GMAO), provides a spatial resolution of approximately 50 × 62.5 km. Its total precipitation and surface temperature data were used. These two reanalysis products are complementary in terms of data assimilation framework and physical parameterization scheme. Average precipitation from both datasets was used to reduce uncertainty and improve the reliability of diagnosing extreme precipitation events and related atmospheric mechanisms.

[0041] Integrated Vertical Water Vapor Transport (IVT): Globally available 3-hour IVT datasets are primarily used to detect Atmospheric Rivers (ARs) events. Based on IVT intensity, ARs events are categorized into five types: AR1 (weak, IVT 250-500 kg / m / s), AR2 (moderate, IVT 500-750 kg / m / s), AR3 (strong, IVT 750-1000 kg / m / s), AR4 (extreme, IVT 1000-1250 kg / m / s), and AR5 (anomaly, IVT > 1250 kg / m / s).

[0042] II. Establish a framework for identifying and attributing AIS quality anomaly events based on the acquired multi-temporal resolution GRACE / FO data (observation data). For example... Figure 1As shown, it includes five main steps: (1) GRACE / FO data processing and uncertainty estimation; (2) significance detection; (3) accumulation period and ablation cycle; and (4) quality anomaly event detection and attribution analysis.

[0043] Step 1: GRACE / FO Data Processing and Uncertainty Estimation. Using the monthly GRACE / FO solution, identify the months from 2002 to 2024 where AIS quality changes are statistically significant. Classify each of these months as either a quality accumulation period or a quality ablation period. This classification enables the diagnosis of whether a significant quality change constitutes a quality anomaly event and extracts the months characterized by quality anomalies.

[0044] The GRACE / FO error assessment was conducted using monthly GRACE / FO mascon solutions from three organizations: CSR, JPL, and GSFC. The specific error assessment details are as follows:

[0045] Mass change time series data from the AIS system were extracted from three mascon datasets, and the monthly gravity monitoring data from different institutions were represented as follows: M c , M j and M g To mitigate the potential model bias inherent in the processing of each dataset, the analysis was performed using the average of three data points:

[0046] ; Where t represents time.

[0047] The triangular cap (TCH) method was applied to assess the uncertainty of these three data points. Assuming the product mass concentration is mascon... It can be represented as a real signal Sum of error terms The sum (the formula is given here using one data point as an example):

[0048] ; Based on this assumption, the difference between observations can be expressed as: ; By calculating the variances on both sides of the above equation and considering the relationship between noise variances, the following variance equation is obtained: ; Assuming the errors among the three Mascon data points are independent, then their covariance... If it is zero, each one can be calculated directly. The variance and root mean square error of the dataset are shown below: ; The uncertainty of each data point for each month was calculated. Using the law of error propagation, the RMS error associated with the monthly average of the three data points was obtained.

[0049] ; Calculate the first difference of the three mascon average time series to represent the monthly quality variation of the AIS system, and perform RMS on the difference time series. Error assessment: ; .

[0050] Step 2: Detection of months with significant quality changes, using a threshold of three times the RMS error to assess the significance of monthly quality changes in the AIS system: ; If the relationship between quality change and error satisfies the above formula, and GRACE / FO observation data for adjacent months are not missing, then the quality change for that month has exceeded the reasonable fluctuation range of observation error, and a significant increase or decrease in quality is considered to have occurred in that month. Conversely, if the quality change is submerged in triple RMS error, it indicates that it may be masked by observation error or uncertainty in data calculation, making it difficult to confirm its statistical significance, and therefore it is not considered a significant change.

[0051] Step 3: Accumulation and Absorption Cycles. The accumulation and absorption cycles of the Antarctic Ice Sheet (AIS) are delineated based on surface mass balance (SMB) data. AIS mass variation is primarily driven by changes in SMB and ice emissions (ID), with basement melting being a secondary contributing factor. Net mass gains or losses in the AIS arise from the imbalance between SMB and ID. However, accurately estimating ID remains challenging. To better characterize the interannual variability of AIS mass variation, all months of each year are classified as either accumulation or absorption periods based solely on SMB data. Between 1979 and 2008, the AIS was roughly in mass balance, meaning that SMB and ID were equal during this period. The accumulation and absorption cycles are then plotted accordingly using this period as a baseline. Specifically, this includes:

[0052] Monthly SMB time series for the entire AIS region from 1979 to 2008 were calculated based on three SMB datasets. The average SMB for all months within this reference period was used as the representative of the ID:

[0053] ; In the formula, SMB refers to any one of the three SMB products, and n is the total number of months in the 30-year reference period from 1979 to 2008 (n = 30 × 12 = 360). Then, the ID is subtracted from the SMB time series to calculate the Quality Anomaly (MA):

[0054] ; Finally, calculate the annual cycle (AC) of AIS quality for each calendar month: ; In the formula, y represents the years from 1979 to 2008, and m represents the calendar month.

[0055] The Accumulation Cycle (AC) is derived from the average monthly Mass Average (MA) over the period from 1979 to 2008. This cycle reflects the typical pattern of mass gain or loss for each calendar month under long-term climatic conditions. A positive phase of the AC for a given month indicates a sustained trend of mass accumulation and is defined as an accumulation period; conversely, a negative phase indicates an ablation period. This classification reflects the typical seasonal response of the AIS system to atmospheric forcing. The core purpose of delineating accumulation and ablation periods is to provide a reference for identifying anomalies in ice sheet mass change events. When a significant mass increase (or loss) event occurs during ablation (or accumulation), it indicates a possible deviation from the normal AC pattern and can be classified as a mass anomaly event. If the trend is consistent with the AC, it is considered a normal fluctuation. This method provides a robust climatic baseline for identifying ice sheet mass change events, enhancing the physical meaning and climatic interpretability of the analysis results.

[0056] Step 4: Anomaly detection and attribution analysis.

[0057] III. Internal verification based on GRACE / FO.

[0058] (1) Changes in AIS quality of GRACE / FO. Compared to the monthly solution, the 10-day and daily GRACE / FO solutions offer higher temporal resolution while maintaining close consistency with the monthly data in terms of long-term trends and annual cycle amplitudes, such as... Figure 2 As shown. Furthermore, all three datasets exhibit a consistent spatial distribution pattern and direction of mass change trends. For example, continuous mass loss is observed along the Amundsen and Beringshausen Sea coastlines, as well as Wilkes Land and Victoria Land, while DronningMaud Land and Enderby Land show a long-term trend of mass increase. This further validates the reliability of the 10-day and daily GRACE / FO solutions in Antarctica.

[0059] Compared to monthly gravity field solutions, daily solutions are more susceptible to the effects of sparse satellite trajectory coverage within a single day. The ITSG daily solution used in this study was derived from Level 1b observations through specific data processing methods, including filtering, regularization, and merging background models, to produce daily Level 2 spherical harmonic coefficients, thus mitigating the limitations of single-day observations. However, in regions with strong gravity signal gradients, such as coastal areas, daily solutions are more prone to signal leakage, which can introduce systematic biases. However, this study focuses on the entire AIS, a large-scale region characterized by continuous landmasses, without involving mixed land-sea grid cells. Therefore, the small differences between daily, 10-day, and monthly solutions can be neglected.

[0060] (2) To ensure a more reliable division between the accumulation and ablation periods, three different SMB data sets and their averages were used. The ID component was removed from the SMB data to highlight interannual anomalous variations in ice sheet mass, such as... Figure 3 As shown in (a), the time series fluctuations of the three data points are highly consistent, exhibiting good synchronicity. To further reduce systematic bias between data sources, the three data points were averaged to construct a more comprehensive composite series. Figure 3 In (b), the annual cycles and averages of the three SMB data sets were calculated. The results show that, although minor differences exist, their intra-annual variation patterns and the months with positive and negative phases are almost identical. The pairwise correlation coefficients for the annual cycles of the three SMB data sets are 0.98, 0.99, and 0.99, respectively. The months with the greatest differences (September and October) coincide with the transition periods simulated by different models with greater differences. Overall, this indicates a high degree of consistency among different climate models in describing the annual cycles of the Antarctic SMB.

[0061] Despite differences in model structure, physical process parameterization, and reanalysis forced data among the three types of SMB data, their high degree of consistency in annual cycle characteristics provides a reliable foundation for accurately identifying AIS quality anomalies in subsequent GRACE / FO quality change analysis. Figure 3 (b) defines the period from March to September as the ice sheet accumulation period. During this phase, temperatures are extremely low, and surface melting is negligible, allowing snowfall to accumulate effectively on the ice sheet surface, driving a net increase in ice sheet mass. Conversely, the period from October to February of the following year is designated as the ice sheet mass ablation period. During this phase, Antarctic temperatures rise significantly, and solar radiation intensifies. Although the overall Antarctic environment remains cold, some areas experience surface melting due to high energy input. Furthermore, wind-driven transport and the intrusion of dry air enhance surface sublimation and evaporation processes, leading to surface mass loss. This dominant mass output results in a mass loss trend in the ice sheet during the summer, thus forming the main interannual rhythm of ice sheet mass balance.

[0062] (3) To identify significant quality change events, a threshold of three times the RMS error is used to select months with statistically significant quality changes. This is achieved by comparing monthly quality changes with three times the RMS error in both positive and negative directions, such as... Figure 4 As shown, significant quality changes were identified over a total of 32 months, with 22 occurring in the GRACE phase and 10 in the GRACE-fo phase.

[0063] (4) Significant changes in quality over 32 months are classified according to whether they occur during the accumulation or ablation phase, such as Figure 5 As shown. The specific classification criteria are as follows: If a significant mass loss event occurs during the accumulation period, or a significant mass increase event occurs during the ablation period, the event will be defined as an anomalous event that deviates from the typical seasonal accumulation-ablation pattern.

[0064] According to this criterion, four anomalous quality events were identified among the 32 months with significant quality changes from 2002 to 2024: anomalous quality loss in September 2003, anomalous quality loss in July 2004, anomalous quality increase in October 2004, and anomalous quality increase in January 2012. These events deviate significantly from the typical response of the AIS system to seasonal climate forcing, suggesting that they may have been influenced by sudden or short-term processes.

[0065] IV. Abnormal Event Analysis.

[0066] (1) Analysis of four abnormal events.

[0067] Reveal the causes of the four identified anomalous events, such as Figure 5 As shown, this is a crucial step in understanding the short-term dynamics of AIS. In terms of driving mechanisms, AIS quality changes are mainly affected by two factors: (1) SMB, which is the sum of precipitation, evapotranspiration, runoff, and sublimation; and (2) ID, which refers to ice mass loss caused by ice dynamics, i.e., the transport and redistribution of ice between the ice sheet and the ocean due to changes in ice flow velocity. Since short-term (e.g., intraday to several days) ID changes tend to be relatively stable and approximately linear, short-term nonlinear quality anomalies on sub-monthly timescales are almost unrelated to ID. Based on this, SMB is the primary focus when analyzing short-term quality anomaly events in the AIS region. In addition, under the unique climatic conditions of Antarctica, the contributions of evapotranspiration, runoff, and sublimation are relatively small (Mottrametal., 2021). Precipitation, as the dominant component of SMB, plays a decisive role in driving AIS quality anomalies. Therefore, precipitation is the primary focus, and its relationship with GRACE / FO-inferred AIS quality changes at different time resolutions (i.e., monthly, 10-day, and daily solutions) during anomaly events is studied.

[0068] For each anomalous event, the corresponding three-month window (the anomalous month plus the preceding month and the following month) contains the GRACE / FO solution and precipitation data as follows: Figure 6 and Figure 7 As shown. It's important to note that the monthly GRACE / FO solution represents the average quality status for the entire month and cannot show changes that occurred on a specific date or within a short period. To assess the reliability of the 10-day and daily solutions during anomalies, their consistency with the monthly solution is evaluated by averaging them on a monthly scale, as shown. Figure 6 As shown. Then, based on the detrended daily solution, the annual cycle of quality change was extracted, and a 5-day fixed average was applied to the daily solution and annual cycle series to suppress high-frequency noise while highlighting low-frequency characteristics, such as... Figure 7 As shown. By comparing the two series, the duration of the impact of each anomalous event can be further estimated. Furthermore, precipitation data were used to characterize changes in surface quality input during the same period. To analyze in detail the relationship between GRACE / FO mass changes and SMB and precipitation changes in each event, Table 1 lists the mass anomalies and corresponding surface quality components for the anomalous months and their adjacent months. Figure 7 Based on the comprehensive analysis in Table 1, a systematic investigation was conducted on the four abnormal events of AIS.

[0069] Table 1. Monthly Grace / FO solutions, precipitation, and SMB quality anomalies for the four anomalous events. Monthly GRACE / FO solutions show an anomalous increase in AIS quality in January 2012. This signal is also clearly present in the monthly averages of the 10-day and daily solutions, confirming that this event can be detected at multiple time resolutions, demonstrating strong reliability and temporal consistency. Figure 6 As shown in (a). Further examination of the daily GRACE / FO solution time series and precipitation data revealed that a five-day heavy precipitation event in December 2011 triggered a positive ice sheet quality anomaly for approximately 50 days. During this period, ice sheet quality was significantly higher than in other years for the same period, such as... Figure 7As shown in (a), based on diurnal solutions, the estimated average positive mass anomaly over the 50-day anomaly period is 134.36 Gt, equivalent to a global mean sea level drop of approximately 0.37 mm. After removing trend and seasonal components, the SMB anomaly and precipitation anomaly increased by 69.49 Gt and 70.14 Gt respectively in December 2011, significantly greater than in other months, as shown in Table 1. In contrast, the SMB anomaly rate and precipitation anomaly rate in January 2012 were only 1.79 Gt and -6.09 Gt respectively, indicating that most of the mass input occurred in December 2011 and did not continue into the following month. This increase in mass anomaly was mainly triggered by the heavy precipitation event in December 2011 and can be classified as a typical short-term, surface input-driven rapid mass increase event of the ice sheet.

[0070] Monthly GRACE / FO solutions revealed a significant quality increase event in October 2004 (Event 2), a signal also clearly captured in daily and 10-day solutions during the same period, such as... Figure 6 As shown in (b). It is worth noting that, according to Figure 3 As shown in b, the annual SMB cycle falls in October, which is a transitional period between the accumulation and dissipation phases, thus lacking clear seasonal representativeness. Therefore, no further detailed study of this event was conducted.

[0071] An anomaly in ice sheet mass loss occurred in July 2004 (Event 3). Further analysis of the daily GRACE / FO solutions indicated that this event could be traced back to mid-June and peaked in July, characterized by a negative ice sheet mass anomaly lasting approximately 50 days. During this period, ice sheet mass was significantly lower than in other years for the same period. Figure 7 (c). According to the daily solution, the average mass anomaly over the approximately 50-day anomalous period is -120.59 Gt, equivalent to a global mean sea level rise of about 0.33 mm. Notably, although all three time resolutions of the GRACE / FO solution show mass loss, precipitation data for the same period do not show any signs of anomalous deficit. As can be seen from Table 1, the average SMB anomaly and precipitation anomaly in June and July 2004 are positive, indicating that surface mass input is in an accumulation phase. The significant difference between surface mass accumulation in regional climate models and mass loss observed by GRACE / FO strongly suggests that this event is not driven by surface processes. Previous studies have shown that ID generally follows a near-linear trend. To isolate short-term anomalies, long-term trend and seasonal components were removed from the GRACE / FO data to exclude the influence of ID. The resulting signals, such as Figure 8 As shown in (a), this reflects the mass anomalies after removing the ID component, including the combined effects of surface and internal glacial processes. Furthermore, in Figure 8As shown in (c), the contribution of SMB was subtracted from the mass anomaly derived from GRACE, leaving a residual signal representing the internal ice mass anomaly. The results indicate that the main area of ​​mass loss has expanded from the coastal edge to a large area within the AIS, suggesting that extensive internal processes are the primary driver of this event. This is highly consistent with the characteristics identified in the time series analysis.

[0072] In contrast, the quality anomaly loss event in September 2003 (Event 4) was relatively simple. For example... Figure 7 As shown in (d), this mass loss event lasted from mid-August to early October 2003, resulting in a negative mass anomaly lasting approximately 50 days, with the most pronounced anomaly occurring in September. Therefore, the monthly solution identifies September as the main month for the anomalous mass loss. During this period, the average mass anomaly was approximately 99.78 Gt, roughly corresponding to a global mean sea level rise of 0.27 mm. Comparison of daily solution time series with precipitation data reveals that these mass loss processes closely coincide with periods of precipitation deficit. Furthermore, as shown in Table 1, the precipitation anomaly for that month was only 1.39 Gt, and the SMB anomaly was negative, indicating a severe deficit in surface mass input. Generally, on short timescales, ID exhibits a linear process. In the context of continuous ID, insufficient surface input naturally leads to ice sheet mass loss. Therefore, the cause of event 4 is ice sheet mass loss due to the precipitation deficit process in that month, occurring against a continuous ID background.

[0073] Analysis of four anomalous events allowed for the identification of the underlying mechanisms and duration of each event. Event 1, caused by a 5-day extreme precipitation event, triggered a positive mass anomaly on the AIS system lasting approximately 50 days, highlighting the rapid response of ice sheet mass to extreme weather events. Event 3 involved a negative mass anomaly of approximately 50 days, primarily driven by non-surface processes. Event 4, characterized by insufficient intra-month precipitation combined with persistent ID (indicating a decrease in surface input), resulted in a negative mass anomaly of approximately 50 days, illustrating a typical case of short-term mass depletion driven by insufficient surface input and persistent ID. These events collectively reveal the multi-source drivers and complexity of AIS mass changes. In addition to surface processes, subglacial and internal dynamic processes also play a crucial role in regulating ice sheet mass changes. Although the daily GRACE / FO solution has limited spatial resolution, its high temporal resolution provides a unique advantage for accurately identifying the timing and evolution of mass anomaly events. Therefore, the daily GRACE / FO solution is an indispensable tool for studying the rapid response mechanisms behind short-term ice sheet mass anomalies.

[0074] 1. Validation based on GNSS displacement time series.

[0075] Based on daily GRACE / FO solutions and precipitation data, it was initially determined that Event 1 was triggered by a 5-day extreme precipitation event, which resulted in a 50-day positive mass anomaly over AIS. To further verify this extreme precipitation event, based on the theory of crustal elastic surface deformation, the vertical displacement time series of 16 GNSS stations were analyzed to investigate whether this event could also be captured in GNSS surface deformation signals.

[0076] Although the data underwent preprocessing, a small number of missing values ​​remained. To construct a complete time series, a composite model incorporating linear trend and annual and semi-annual periodic interpolation was used to fill in the gaps. Before filling in the missing values, it was ensured that no missing values ​​occurred during periods of extreme precipitation to avoid affecting the detection of key signals. Figure 9 The red curve in the blue shaded area of ​​(b). Examining the spatial distribution of precipitation reveals that this extreme precipitation event was primarily concentrated in the Antarctic Peninsula, Dronning Maud Land, Wilkes Land, and the Amundsen Sea coastline, such as... Figure 9 (a) Due to varying local environmental conditions, system effects may influence the vertical displacement of GNSS. To mitigate these effects, for small areas containing multiple GNSS stations, Figure 9 In regions I and II of (a), principal component analysis (PCA) was applied to extract and remove common mode error (CME). The time series was refitted and the trend and seasonal components were removed to obtain a clearer vertical displacement residual signal. Then, a 5-day moving average was applied to smooth the residual series, effectively filtering out high-frequency noise while retaining low-frequency trend features.

[0077] Analysis of the GRACE / FO diurnal solution confirmed that the extreme precipitation event triggered a significant positive mass anomaly lasting approximately 50 days, such as... Figure 7 As shown in (a). To further verify this result, vertical displacement data from 16 GNSS stations located in four precipitation concentration areas were analyzed, and data from the same period in two adjacent years (December 16, 2010 - March 1, 2011 and December 16, 2011 - March 1, 2012) were compared. The results show that within 50-70 days after a 5-day extreme precipitation event, the vertical displacement of almost all GNSS stations was lower than that of the same period in the previous year. Figure 9In (b), the red shading area is larger than the green shading area, indicating that the GNSS observations successfully captured the response to the event. Although some stations (such as OHI3 and DUM1) generally support the signal of the event, they show a distinct green area, indicating that GNSS measurements inevitably contain superimposed complex geophysical signals and observational noise. Despite comprehensive corrections and processing, external factors may still have some influence. Nevertheless, the overall results from all stations clearly demonstrate that the GNSS vertical displacement observations successfully captured this extreme precipitation event.

[0078] 2. Atmospheric driving factors of extreme precipitation events.

[0079] From December 16 to 20, 2011, the AIS region experienced five consecutive days of heavy precipitation, forming a 50-day period of positive mass anomalies. According to the Oceanic Niño Index (ONI) of the National Oceanic and Atmospheric Administration (NOAA), the value in December 2011 was -1.04°C, indicating a moderate LaNiña event (between -1°C and -1.5°C) (Wang et al., 2025). Meanwhile, the South Annular Mode (SAM) index provided by the British Antarctic Survey (BAS) reached +3.43 in December 2011 and remained high at +3.08 in January 2012, both indicating strong positive phases. Although ENSO and SAM conditions did not directly trigger extreme precipitation, they together established a favorable large-scale climate background. The LaNiña event enhanced convection activity over the western Pacific warm pool, indirectly triggering anomalies in the mid-latitude circulation of the Southern Hemisphere. Meanwhile, the strong positive SAM phase causes the westerly jet stream to shift southward, which intensifies the interaction between cold and warm air masses and enhances meridional water vapor transport from mid-latitudes to the poles. Therefore, subsequent synoptic-scale circulation evolution and water vapor transport processes can develop under this large-scale climate setting, ultimately leading to extreme precipitation events.

[0080] The 500 hPa geopotential height field shows four quasi-symmetrically distributed low-pressure centers over the Southern Ocean, such as... Figure 10 These systems (ae) remained spatially stable over a 5-day period, with extreme precipitation concentrated primarily near the high-pressure ridges between them, indicating that upper-level wave structure plays a crucial role in guiding water vapor convergence and upward motion. The 850 hPa specific humidity field exhibits strong spatial coherence between high-humidity areas and coastal precipitation belts, particularly along the Antarctic Peninsula, Dronning Maud Land, George V Land, and the Amundsen Sea, such as... Figure 10(fj). These high-humidity areas are formed by the continuous advection of the westerly winds and subtropical jet stream from the mid-latitudes of the southern Indian Ocean and the South Pacific. Their movement is closely aligned with the wind direction, indicating that the continuous rise of moist air sustains heavy precipitation. Simultaneously, the sea-level pressure field contains four surface low-pressure systems, located very close to the 500 hPa low-pressure system, exhibiting a vertically aligned circulation structure, such as... Figure 10 The vertical consistency between the surface and upper-level systems provides sustained dynamic support for water vapor accumulation and deep convection, underpinning the development and persistence of extreme precipitation events.

[0081] Furthermore, analysis of IVT data before and after precipitation revealed a close link between extreme precipitation events and ARs. December 16th marked the beginning of heavy precipitation, such as... Figure 11 As shown in (a), the heavy precipitation in the Antarctic Peninsula and Dronning Maud Land spatially coincides with a significant IVT transport path originating from the mid-latitudes (30°S–40°S) in the early stages. To better characterize the spatiotemporal development of ARs, further investigation was conducted. Figure 11 The IVT evolution is shown in the figure. The results indicate that ARs first made landfall over the Antarctic Peninsula at 12:00 on December 15th and reached Dronning Maud Land at 21:00 on the same day. The AR intensity reached its peak over the Antarctic Peninsula at 9:00 on December 16th, reaching AR level 3 (strong); and reached its maximum intensity over Dronning Maud Land at 00:00 on the 17th, reaching AR level 2 (moderate). The spatiotemporal evolution of ARs shows a high degree of consistency with precipitation events, indicating that ARs played a crucial role in transporting abundant water vapor from the mid-latitudes to the Antarctic coast, significantly enhancing precipitation processes. Combined with... Figure 10 The dynamic conditions shown indicate that water vapor carried by ARs was lifted and condensed along the coastal topography, ultimately triggering a sustained and intense extreme precipitation event. ARs not only complement the large-scale climate and weather dynamics discussed earlier, but also highlight the direct water transport mechanism that triggered and sustained this extreme precipitation event.

[0082] The formation and persistence of the five-day extreme precipitation event in Antarctica in December 2011 were driven by a series of factors: a moderate-intensity La Niña event and a strong positive phase providing a favorable climate background; a stable and symmetrically distributed low-pressure system over the Southern Ocean; a coherent vertical atmospheric structure; and the occurrence of an ARs (Air-Average Reduction) event. ARs act as an important channel, transporting abundant moisture from mid-latitude regions to the Antarctic coast, thereby enhancing the moisture supply required for sustained heavy precipitation. This event illustrates the coupling mechanism between large-scale climate variability, weather circulation patterns, water vapor transport pathways including ARs, and vertical atmospheric dynamics.

[0083] This invention also provides a system for identifying anomaly events in the Antarctic ice sheet, comprising: The data module is used to acquire GRACE / FO gravity data within a set time period; the set time period is measured on a monthly, 10-day, or daily scale.

[0084] The anomaly identification module is used to acquire the surface mass balance (SMB) data of the Antarctic ice sheet, and to divide the Antarctic ice sheet AIS into accumulation and ablation periods based on the SMB data; it performs monthly mass change error assessment based on GRACE / FO gravity data, sets a significance threshold based on the monthly mass change error assessment results, and marks months exceeding the significance threshold as months with significant mass changes; if a month with significant mass changes experiences mass loss during the accumulation period or mass increase during the ablation period, it is judged as an Antarctic ice sheet mass anomaly event.

[0085] The modules in the aforementioned Antarctic ice sheet quality anomaly event identification system can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the computer device's memory as software, so that the processor can call and execute the corresponding operations of each module.

[0086] The present invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory. The processor executes the computer program to implement the steps in an embodiment of a method for identifying anomalies in Antarctic ice sheet mass. Specific implementation methods can be found in the method embodiments, and will not be repeated here.

[0087] Furthermore, the present invention also provides a non-transitory computer-readable storage medium containing instructions on which a computer program is stored. For example, a memory containing instructions that can be executed by a processor of a computer device to perform the above-described method. For example, the non-transitory computer-readable storage medium may be a ROM, random access memory (RAM), CD-ROM, magnetic tape, floppy disk, and optical data storage device, etc. When the computer program is executed by the processor, it can implement the steps in an embodiment of a method for identifying anomalies in Antarctic ice sheet quality. Specific implementation methods can be found in the method embodiments, which will not be repeated here.

[0088] Those skilled in the art will understand that embodiments of the present invention can provide methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0089] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, as well as combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0090] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0091] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0092] It should be noted that the specific embodiments described above enable those skilled in the art to more fully understand the present invention, but do not limit the present invention in any way. Therefore, although the present invention has been described in detail in this specification and embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the present invention; and all technical solutions and improvements that do not depart from the spirit and scope of the present invention are covered within the protection scope of the present invention patent. No reference numerals in the claims should be construed as limiting the scope of the claims. Any simple variations or equivalent substitutions of technical solutions that can be readily obtained by those skilled in the art within the scope of the technology disclosed in the present invention are within the protection scope of the present invention.

Claims

1. A method for identifying anomaly events in the Antarctic ice sheet, characterized in that, Includes the following steps: Acquire GRACE / FO gravity data within a specified time period; the specified time period is measured on a monthly, 10-day, or daily scale. Obtain surface mass balance (SMB) data of the Antarctic ice sheet, and divide the accumulation and ablation periods of the Antarctic ice sheet AIS based on the SMB data; Monthly mass change error assessment is performed based on the GRACE / FO gravity data. A significance threshold is set based on the monthly mass change error assessment results, and months exceeding the significance threshold are marked as months with significant mass changes. If a month with significant mass changes experiences mass loss during the accumulation period or mass increase during the melting period, it is judged as an Antarctic ice sheet mass anomaly event.

2. The method for identifying anomaly events in the Antarctic ice sheet according to claim 1, characterized in that, The division of the accumulation and ablation periods of the Antarctic ice sheet AIS based on the aforementioned SMB data includes the following steps: Use the average of the SMB time series data for all months within a specified time period as the ID data; The quality anomaly data is obtained by subtracting the ID data from the SMB time series of all months within a set time period; Based on the aforementioned quality anomaly data, the annual cycle AC of AIS quality is calculated. If the monthly AC is positive, it is the accumulation period; if it is negative, it is the ablation period.

3. The method for identifying anomaly events in the Antarctic ice sheet according to claim 1, characterized in that, The monthly mass change error assessment based on the GRACE / FO gravity data includes the following steps: Calculate the average monthly GRACE / FO gravity data from different datasets; wherein, the different datasets are monthly gravity monitoring data from different institutions; The variance between different datasets is calculated using the triangular hat method, and the RMS error is calculated based on the variance. The first difference of the monthly average series is obtained to obtain the monthly quality change difference series and the corresponding RMS error for each month.

4. The method for identifying anomaly events in the Antarctic ice sheet according to claim 1, characterized in that, The significance threshold is specifically: ； in, Monthly quality changes for the AIS system. This is the RMS error; If the relationship between quality change and error satisfies the above formula, and the GRACE / FO observation data for adjacent months are not missing, then a significant increase or decrease in quality is considered to exist in that month; conversely, if the quality change is submerged in 3 times the RMS error, it is not considered a significant change.

5. The method for identifying anomaly events in the Antarctic ice sheet according to claim 1, characterized in that, It also includes determining the driving proportion of the results of the aforementioned quality anomaly events, specifically including the following steps: The original time series of AIS mass changes is extracted from the GRACE / FO gravity data; the long-term trend, annual period and semi-annual period of the original time series are fitted to obtain the residual series after deducting the trend term and period term from the original mass time series, and the difference value is calculated. The contribution ratio is calculated based on the difference value of the original quality time series and the quality difference value after removing the trend and periodic components. If the contribution ratio is positive, it means that the quality change becomes more obvious after removing the long-term trend and periodic fluctuations, and the quality change is dominated by short-term processes. If the contribution ratio is negative, it means that the trend and periodic components have explanatory power for the quality change of the month, and the quality change is regarded as the result of long-term climate forcing.

6. A system for identifying anomaly events in the Antarctic ice sheet, characterized in that, include: The data module is used to acquire GRACE / FO gravity data within a specified time period. The specified time period is measured in monthly, 10-day, and daily increments. An anomaly identification module is used to acquire the surface mass balance (SMB) data of the Antarctic ice sheet, and to divide the Antarctic ice sheet AIS into accumulation and ablation periods based on the SMB data; to perform monthly mass change error assessment based on the GRACE / FO gravity data, and to set a significance threshold based on the monthly mass change error assessment results, marking months exceeding the significance threshold as months with significant mass changes; if a month with significant mass changes experiences mass loss during the accumulation period or mass increase during the ablation period, it is judged as an Antarctic ice sheet mass anomaly event.

7. A computer device, comprising a memory, a processor, and a computer program stored in the memory, characterized in that, The processor executes the computer program to implement the steps of the method according to any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is loaded by the processor, it is able to perform the steps of the method according to any one of claims 1 to 5.

Images