Remote sensing of lake ice phenology in Alaska
Remote sensing of lake ice phenology in Alaska
复制标题
DOI：
10.1088/1748-9326/abf965
复制
发表时间：
2021-05
影响因子：
6.7
通讯作者：
Shuai Zhang;T. Pavelsky;C. Arp;X. Yang
中科院分区：
文献类型：
--
作者：
Shuai Zhang;T. Pavelsky;C. Arp;X. Yang
湖冰破裂和冻结的时间是北极和北方地区气候变化的重要指标，因为它们对气候条件的变化作出迅速和直接的反应。尽管其重要性，湖冰物候仍然很少记录在阿拉斯加的大多数湖泊。为了填补这一数据空白，我们构建了一个遥感衍生的湖冰物候数据库，涵盖了2000-2019年期间阿拉斯加所有大于1平方公里（n = 4241）的湖泊。该数据集，其中包括湖冰开/关日期和湖冰持续时间，是基于一个自动的方法，每天使用中分辨率成像光谱仪（MODIS）图像来测量湖冰分数。该方法利用Landsat Fmask标定的动态阈值从MODIS图像中提取湖冰像素。不同的过滤器，占云，极夜，和其他来源的误差，以提高湖冰物候估计的准确性。趋势分析显示，440个湖泊的解体较早（-5.5 ddecade-1），四湖的解体较晚（7.5 ddecade-1）（p < 0.05）。共有289个湖泊有明显的封冻后期趋势（2.9 d decade−1），11个湖泊有封冻早期趋势（-3.3 d decade−1）。大多数具有显著趋势的湖泊位于布鲁克斯山脉以北。该数据集有助于加深对湖泊过程与气候变化之间相互作用的理解，并支持对阿拉斯加和泛北极地区的地球化学、湖沼学和生态制度的研究。
The timing of lake ice breakup and freezeup are important indicators of climate change in Arctic and boreal regions because they respond rapidly and directly to variations in climate conditions. Despite its importance, lake ice phenology remains poorly documented in most lakes of Alaska. To fill this data gap, we constructed a remote sensing-derived lake ice phenology database covering all lakes in Alaska larger than 1 km2 (n = 4241) over the period 2000–2019. This dataset, which includes lake ice on/off dates and lake ice duration, was based on an automatic method using daily moderate resolution imaging spectroradiomenter (MODIS) imagery to measure lake ice fraction. This method extracts lake ice pixels from MODIS images using a dynamic threshold that was calibrated against Landsat Fmask. Different filters that account for clouds, polar night, and other sources of error were applied to increase the accuracy of lake ice phenology estimation. Trend analysis shows earlier breakup (−5.5 d decade−1) for 440 lakes and later breakup (7.5 d decade−1) for four lakes (p < 0.05). A total of 289 lakes had significant trends toward later freezeup (2.9 d decade−1) and 11 lakes towards earlier freezeup (−3.3 d decade−1). Most lakes with significant trends are north of the Brooks Range. This dataset can contribute to increased understanding of interactions between lake processes and climate change, and it supports the study of biogeochemical, limnological and ecological regimes in Alaska and pan-Arctic regions.