北半球湖冰厚度的现状及未来

The state and fate of lake ice thickness in the Northern Hemisphere

湖冰厚度是对气候变化高度敏感的指示因子,通过影响湖泊与大气之间的物质能量交换,调节区域水文气候和湖泊生态系统,并对维系冰上生产生活至关重要.由于实测湖冰厚度资料极度匮乏,同时缺乏高效的遥感观测手段,大范围湖冰厚度的时空演变规律还存在空白.现有的全球湖冰数据主要利用湖冰模型基于再分析气象驱动场和虚拟湖泊生成,难以反映真实湖冰厚度.本研究使用卫星雷达测高反演了北半球16个大型湖泊和水库过去近30年的冰厚,精度约为0.2m.在此基础上开发并验证了遥感数据驱动的湖冰模型,模拟了北半球大于50 km^(2)的1313个湖泊和水库(湖库总面积约840,000 km^(2))的冰厚,估算了2003~2018年北半球多年平均最大冰厚为0.63 m,相当于485 Gt水当量.以该时段冰厚为参考,预估了在RCP 8.5情境下未来(2071~2099年)最大冰厚可能减小约0.35 m,北半球湖冰变薄将对区域水文气候和湖泊生态系统,以及冰上交通等社会经济活动造成重大影响.

Lake ice thickness(LIT)is important for regional hydroclimate systems,lake ecosystems,and human activities on the ice,and is thought to be highly susceptible to global warming.However,the spatiotemporal variability in LIT is largely unknown due to the difficulty in deriving in situ measurements and the lack of an effective remote sensing platform.Despite intensive development and applications of lake ice models driven by general circulation model output,evaluation of the global LIT is mostly based on assumed“ideal”lakes in each grid cell of the climate forcing data.A method for calculating the actual global LIT is therefore urgently needed.Here we use satellite altimetry to retrieve ice thickness for 16 large lakes in the Northern Hemisphere(Lake Baikal,Great Slave Lake,and others)with an accuracy of~0.2 m for almost three decades.We then develop a 1-D lake ice model driven primarily by remotely sensed data and cross-validated with the altimetric LIT to provide a robust means of estimating LIT for lakes larger than 50 km^(2)across the Northern Hemisphere.Mean LIT(annual maximum ice thickness)for 1313 simulated lakes and reservoirs covering~840,000 km^(2)for 2003–2018 is 0.63±0.02 m,corresponding to~485 Gt of water.LIT changes are projected for 2071–2099 under RCPs 2.6,6.0,and 8.5,showing that the mean LIT could decrease by~0.35 m under the worst concentration pathway and the associated lower ice road availability could have a significant impact on socio-economic activities.