摘要
近地表土壤冻融循环过程是陆面过程的"开关",对地表能量过程、水文过程、植被动态、温室气体交换和生态系统功能等都具有重要影响。除了依靠传统观测手段判别地表冻融,被动微波卫星遥感技术已被成熟运用于全球尺度地表冻融研究中。利用被动微波卫星遥感亮温监测地表冻融大致经历了理论基础和算法的研制、算法参数的验证和应用以及建立数据集三个主要过程,并基于不同的算法针对不同目的分别发布了不同地区、不同时间序列的近地表土壤冻融数据集。以中国气象局435个气象台站观测的0cm土壤温度为标准,对比分析了基于目前较为通用的双指标算法、单指标算法、决策树算法和判别式算法建立的近地表土壤冻融数据集的分类结果。结果表明:单指标算法数据集冻结指标的平均值与观测资料最接近,其偏差和均方根误差最小,而决策树算法由于全局采用单一的37GHz垂直极化亮温阈值,低估了地表冻结,导致偏差和均方根误差最大。近地表土壤冻融数据集和观测资料均表明随着全球变暖,近地表土壤呈冻结首日推迟、冻结终日提前和冻结天数减少的趋势。通过对不同冻融数据集的比较研究,为未来被动微波卫星遥感算法改进、近地表土壤冻融数据的合理使用提供了可靠的科学依据。
The near-surface soil freeze/thaw(F/T)cycle is characterized as a toggle switch for land surface process because of its strong influence on surface energy balance,hydrological processes,vegetation dynamics,greenhouse gas exchange,and ecosystem function.In addition to the traditional observation methods,passive microwave satellite remote sensing technology has been applied to the research of the near surface soil freeze/thaw status on a global scale.The development of detecting near-surface soil F/T status algorithm has experienced three stages:algorithm development,parameterization/validation,and F/T dataset establishment.Based on different algorithms,the near-surface soil F/T status datasets were developed in different regions during different periods.This study aims to compare the classification results of 3 near-surface soil F/T status datasets established based on single-index algorithm(SIA),double index algorithm(DIA)and the decision-tree algorithm(DTA).Soil temperatures at 0 cm depth(Tg)from 435 meteorological stations across China were used as a reference.Our preliminary results indicate that the SIA’s classification results show best agreement with ground observations.DTA underestimated the near-surface soil frozen status because of utilizing the unique 37 GHz threshold of vertically polarized brightness temperature throughout the whole classification process.Datasets and ground-based measurements of the near-surface soil F/T status show that due to the global warming,the onset of soil frozen for near-surface soil delayed,while the offset of soil frozen advanced.The actual number of frozen days had been decreasing in China.This study compared the different datasets of F/T,and provided a scientific basis for the future improvement and utilization of the satellite passive microwave remote sensing data.
作者
邵婉婉
张廷军
SHAO Wanwan;ZHANG Tingjun(Key Laboratory of Western China’s Environmental Systems(Ministry of Education),College of Earth and Environmental Sciences,Lanzhou University,Lanzhou 730000,China)
出处
《冰川冻土》
CSCD
北大核心
2021年第1期285-295,共11页
Journal of Glaciology and Geocryology
基金
中国科学院A类先导科技专项(XDA20100103,XDA20100313)资助。
关键词
近地表土壤冻融
双指标算法
决策树算法
土壤冻融数据集
near-surface soil freeze/thaw
dual index algorithm
decision tree algorithm
soil freeze/thaw status datasets
作者简介
邵婉婉,博士研究生,主要从事冻土研究.E-mail:shaoww09@lzu.edu.cn;通信作者:张廷军,教授,主要从事冰冻圈与气候变化研究.E-mail:tjzhang@lzu.edu.cn。
