近年来，格陵兰冰盖北部地区物质损失加速，冰面消融产生大量融水，融水在地形排水作用下汇集形成冰面融水径流。在溢出冰川以外地区，竖井和注水冰裂隙较少分布，冰面融水被直接汇流至冰前区域形成冰前水系，最终汇入海洋，形成独特的冰面-冰前融水汇流过程，这对冰盖物质平衡以及海洋环境变化产生重要影响。卫星遥感能够直接观测冰面融水径流和冰前水系的时空分布，提供河流位置、形态、动态变化等关键信息，已成为研究格陵兰冰盖融水汇流过程的重要手段。本研究以格陵兰冰盖北部地区Denmark冰面-冰前流域（面积3240 km2）作为研究区，采用Sentinel-2和Landsat-8卫星影像提取了研究区2014-2021年消融期（6-8月）冰面融水范围与流域出口冰前河宽，分析了冰面融水与冰前河的季节与年际变化特征。进一步，对比了遥感观测的冰面-冰前流域融水与区域气候模型（MARv3.12与RACMO2.3p2）模拟的冰面融水径流量，揭示了冰面-冰前融水汇流过程对冰面消融强度的响应。研究结果表明：（1）消融期内，冰面融水范围首先向高海拔地区推进（最高海拔至~1400 m），随后逐步消退至冰盖边缘（至~500 m），流域出口冰前河宽呈现先增大（至~2000 m）后减小（至~100 m）的变化趋势；（2）遥感观测的冰面融水与流域出口冰前河宽呈现显著线性正相关关系(r = 0.87，p < 0.01)，流域内形成了连续的冰面-冰前水文系统，能够有效汇流融水离开冰盖进入海洋；（3）MARv3.12与RACMO2.3p2模型能够较准确地模拟冰面融水径流量，冰面融水径流量与遥感观测的冰面融水（MAR：r = 0.87；RACMO：r = 0.84，p < 0.01）以及流域出口冰前河宽（MAR：r = 0.89；RACMO：r = 0.88，p < 0.01）均具有较强的相关性。（4）考虑融水汇流滞时的冰面融水径流量与流域出口冰前河宽的相关系数（MAR：r = 0.94；RACMO：r = 0.92，p < 0.01）提升，显著高于瞬时冰面融水径流量对应的相关系数，Denmark流域冰面-冰前融水汇流过程的最优滞时约为2天，这一滞时定量表征了Denmark流域冰面-冰前流域输送融水的效率。

Mass loss from the Greenland Ice Sheet (GrIS) has accelerated in recent decades with profound effects on global sea-level rise. During each summer, the meltwater forms supraglacial rivers and then is transported to the proglacial zone, eventually flowing into the ocean and forming a continuous supraglacial-proglacial river system. This continuous supraglacial-proglacial drainage system directly results the mass loss of the GrIS and has an important impact on the changes in the marine environment. Satellite images can directly observe the temporal and spatial distribution of supraglacial and proglacial rivers, and have been widely used in the study of the GrIS. The satellite-derived observation can provide key information such as the location, morphology, and dynamic changes of rivers. It has become an important way to analyze the meltwater routing process. In this paper, 361 scenes of Sentinel-2 and Landsat-8 satellite images are used to extract the supraglacial and proglacial rivers in the Denmark supraglacial-proglacial basin of the northeastern GrIS during the melt seasons (July to August) and monitor their spatial distribution and dynamic changes. Further, satellite-derived observation and meltwater runoff simulated by regional climate models (MARv3.12 and RACMO2.3p2) were compared and analyzed, then estimated the lag time of the supraglacial-proglacial drainage system. The main research contents and conclusions of this paper include the following three aspects: (1) the proglacial river width is in the range of 100 to 2000 m and experiences a seasonal trend. The ice surface meltwater showed similar variation characteristics, advancing to the high-altitude areas of the ice surface (up to ~1400 m) at the initial stage of ablation, and then gradually receding to the edge of the ice sheet (up to ~500 m). (2) there is a significant positive correlation between satellite-derived proglacial river width and meltwater on the ice surface (r = 0.87, p < 0.01), forming a continuous supraglacial-proglacial drainage system which can effectively transport the meltwater each summer. (3) MARv3.12 and RACMO2.3p2 models can accurately simulate the meltwater runoff in the supraglacial-proglacial drainage system on a large area and long-term scale and the simulated meltwater runoff and satellite-derived ice surface meltwater (MAR: r = 0.87; RACMO: r = 0.84, p < 0.01) and proglacial river width (MAR: r = 0.89; RACMO: r = 0.88, p < 0.01) both have strong correlations. (4) the r between the simulated lagged meltwater runoff and satellite-derived proglacial river width (MAR: r = 0.93; RACMO: r = 0.92, p < 0.01) increased, which was significantly higher than that of the instantaneous meltwater runoff. The optimal lag time of the supraglacial-proglacial drainage system in Denmark Basin was about 2 days. This lag time quantitatively represents the efficiency of meltwater routing in the supraglacial-proglacial drainage system.