X img





作者投稿 专家审稿 编辑办公 编委办公 主编办公

冰川冻土 ›› 2019, Vol. 41 ›› Issue (5): 1183-1191.doi: 10.7522/j.issn.1000-0240.2019.1180

• 积雪与冰冻圈遥感 • 上一篇    下一篇


郭慧1, 陈思勇1, 王晓艳1,2   

  1. 1. 兰州大学 资源环境学院, 甘肃 兰州 730000;
    2. 中国科学院 地理科学与资源研究所 资源与环境信息系统国家重点实验室, 北京 100101
  • 收稿日期:2019-07-21 修回日期:2019-09-04 发布日期:2020-02-24
  • 通讯作者: 王晓艳,E-mail:wangxiaoy@lzu.edu.cn. E-mail:wangxiaoy@lzu.edu.cn
  • 作者简介:郭慧(1994-),女,山西大同人,2017年在西安科技大学获学士学位,现为兰州大学在读硕士研究生,从事积雪遥感研究.E-mail:hguo18@lzu.edu.cn
  • 基金资助:

Estimation of fraction of snow cover in Northeast China based on MODIS data

GUO Hui1, CHEN Siyong1, WANG Xiaoyan1,2   

  1. 1. College of Earth and Environmental Sciences, Lanzhou University, Lanzhou 730000, China;
    2. State Key Laboratory of Resources and Environmental Information System, Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing 100101, China
  • Received:2019-07-21 Revised:2019-09-04 Published:2020-02-24

摘要: 东北地区是我国三大积雪区之一,森林覆盖面积占总面积的40%左右。受森林冠层的影响,当前的MODIS雪盖产品(V6)提供的积雪覆盖率标准模型对东北地区积雪覆盖率估算结果存在明显的低估现象。基于此,采用分区建模的方式:在森林地区,计算归一化差值林地积雪指数(NDFSI),建立像元积雪覆盖率(FSC)与NDFSI及NDVI之间的线性关系;在非森林地区,采用MOD10A1 V6提供的归一化差值积雪指数(NDSI),建立像元积雪覆盖率(FSC)与NDSI及NDVI之间的线性关系。采用Landsat 8 OLI数据提取的积雪覆盖率(FSC)对分区建模的估算结果与标准模型的估算结果进行对比,发现进行估算的过程中均方根误差和平均绝对误差这两项指标的数值相对于标准模型有了大幅下降,这一结果在林区有更显著的表现。计算得到的决定系数R2,在本文模型也有提高。以T1林区影像为例,本文模型的均方根误差和平均绝对误差分别为0.246、0.055,而标准模型的两项指标则分别为0.420、0.348。本文模型和标准模型的决定系数分别为0.675、0.641。

关键词: 东北地区, 积雪覆盖率(FSC), MODIS, 分区模型

Abstract: Northeast China is one of the three major snow-covered areas in China, with forest coverage accounting for about 40% of the total area. Due to the influence of forest canopy, the standard model of snow cover provided by MODIS snow cover products (V6) is obviously underestimated. Based on this, the method of zoning modeling was adopted:in the forest area, the normalized difference forest snow cover index (NDFSI) was calculated, and the linear model between fraction of snow cover (FSC) and NDFSI and normalized difference vegetation index (NDVI) was established. In the non-forest areas, NDSI (normalized difference snow cover index) provided by MOD10A1 V6 was adopted to establish a linear relationship between pixel snow cover and NDSI and NDVI. In view of this situation, in this paper, the method of zonal modeling is adopted to calculate the NDFSI in forest areas; in the non-forest areas, the linear relationship between pixel snow cover and NDSI and NDVI is established by using NDSI provided by MOD10A1 V6. The fraction of snow cover extracted by Landsat 8 OLI was used to compare the estimation results of the zoning model with those of the standard algorithm model. The comparison shows that the root mean square error (RMSE) and the mean absolute error (MAE) in the process of estimating have greatly decreased as compared with the standard algorithm model, and this result has a more significant performance in the forest area. The coefficient of determination R2 calculated in this paper has also been greatly improved. Taking T1 forest areas image as an example, the mean square root error and the mean absolute error of the model are 0.246 and 0.055, respectively, while the two indexes of the standard algorithm model are 0.420 and 0.348, and the coefficients of determination of the algorithm model in this paper and the standard algorithm model are 0.675 and 0.641, respectively.

Key words: Northeast China, fraction of snow cover (FSC), MODIS, zoning model


  • TP79