A deep learning reconstruction of mass balance series for all glaciers in the French Alps: 1967-2015

Glacier surface mass balance (SMB) data are crucial to understand and quantify the regional effects of climate on<br> glaciers and the high-mountain water cycle, yet observations cover only a small fraction of glaciers in the world. We present<br> a dataset of annual glacier-wide surface mass balance of all the glaciers in the French Alps for the 1967-2015 period. This<br> dataset has been reconstructed using deep learning (i.e. a deep artificial neural network), based on direct and remote sensing<br> SMB observations, meteorological reanalyses and topographical data from glacier inventories. This data science reconstruction<br> approach is embedded as a SMB component of the open-source ALpine Parameterized Glacier Model (ALPGM: https://zenodo.org/record/3609136). An extensive cross-validation allowed to assess the method’s validity, with an estimated average error (RMSE) of 0.49 m.w.e. a^-¹, an explained variance (r²) of 79% and an average bias of +0.017 m.w.e. a^-¹. We estimate an average regional area-weighted glacier-wide SMB of -0.72±0.20 m.w.e. a^-¹ for the 1967-2015 period, with moderately negative mass balances in the 1970s (-0.52 m.w.e. a^-¹) and 1980s (-0.12 m.w.e. a^-¹), and an increasing negative trend from the 1990s onwards, up to -1.39<br> m.w.e. a^-¹ in the 2010s. Following a topographical and regional analysis, we estimate that the massifs with the highest mass losses for this period are the Chablais (-0.90 m.w.e. a^-¹) and Ubaye and Champsaur (-0.91 m.w.e. a^-¹ both) ranges, and the ones presenting the lowest mass losses are the Mont-Blanc and Oisans ranges (-0.74 and -0.78 m.w.e. a^-¹ respectively). This dataset provides relevant and timely data for studies in the fields of glaciology, hydrology and ecology in the French Alps, in need of regional or glacier-specific meltwater contributions in glacierized catchments. The SMB dataset is comprised of multiple <em>CSV</em> files, one for each of the 661 glaciers from the 2003 glacier inventory (Gardent et al., 2014), named with its GLIMS ID and RGI ID with the following format: <em>GLIMS-ID_RGI-ID_SMB.csv</em>. Both indexes are used since some glaciers that split into multiple sub-glaciers do not have an RGI ID. Split glaciers have the GLIMS ID of their "parent" glacier and an RGI ID equal to 0. Every file contains one column for the year number between 1967 and 2015 and another column for the annual glacier-wide SMB time series. Glaciers with remote sensing-derived observations (Rabatel et al., 2016) include this information as an additional column. This allows the user to choose the source of data, with remote sensing data having lower uncertainties (0.35±0.06 () m.w.e. a^-¹ as estimated in Rabatel et al. (2016)). Columns are separated by semicolon (;). All topographical data for the 661 glaciers can be found in the updated version of the 2003 glacier inventory included in the Supplementary material.冰川表面物质平衡（Surface Mass Balance, SMB）数据对于理解并量化气候对冰川及高山区水循环的区域影响至关重要，但当前观测仅覆盖全球极小一部分冰川。我们构建了1967-2015年期间法国阿尔卑斯山区所有冰川的年度全域表面物质平衡数据集。该数据集基于直接观测与遥感 SMB 观测、气象再分析数据以及冰川目录地形数据，通过深度学习（即深度人工神经网络）完成重建。本数据科学重建方法被嵌入至开源阿尔卑斯参数化冰川模型（ALpine Parameterized Glacier Model, ALPGM：https://zenodo.org/record/3609136）的SMB组件中。 研究通过大量交叉验证评估了该方法的有效性，估算得到平均误差（均方根误差, RMSE）为0.49 米水当量·年⁻¹，解释方差（决定系数r²）为79%，平均偏差为+0.017 米水当量·年⁻¹。我们估算得到1967-2015年法国阿尔卑斯山区区域面积加权的全域冰川SMB为-0.72±0.20 米水当量·年⁻¹，其中20世纪70年代物质平衡为中度负值（-0.52 米水当量·年⁻¹），80年代为-0.12 米水当量·年⁻¹，自90年代起负向趋势加剧，2010年代达到-1.39 米水当量·年⁻¹。 经地形与区域分析，我们估算得到该时期物质损失最严重的山地为夏布莱（Chablais）山脉（-0.90 米水当量·年⁻¹）以及于拜（Ubaye）和尚普索尔（Champsaur）山脉（二者均为-0.91 米水当量·年⁻¹）；物质损失最小的为勃朗峰（Mont-Blanc）和瓦桑（Oisans）山脉（分别为-0.74和-0.78 米水当量·年⁻¹）。 本数据集可为法国阿尔卑斯山区冰川学、水文学与生态学领域的研究提供及时且有价值的数据，这类研究往往需要冰川化流域的区域或特定冰川融水贡献数据。该SMB数据集包含多个逗号分隔值文件（Comma-Separated Values, CSV），对应2003年冰川目录（Gardent等，2014）中的661条冰川，每个文件以其GLIMS ID和RGI ID命名，格式为：GLIMS-ID_RGI-ID_SMB.csv。由于部分冰川分裂为多条子冰川后未分配RGI ID，因此同时使用两种索引。分裂冰川的GLIMS ID保留其“母冰川”的ID，且RGI ID设为0。每个文件包含两列：一列对应1967-2015年的年份，另一列对应年度全域冰川SMB时间序列。带有遥感观测数据（Rabatel等，2016）的冰川会额外增加一列遥感观测数据，方便用户选择数据来源，其中遥感数据的不确定性更低，据Rabatel等（2016）估算为0.35±0.06 米水当量·年⁻¹。列之间以分号（;）分隔。661条冰川的所有地形数据可在补充材料附带的2003年冰川目录更新版本中获取。