Login   |      Register
English    中文


Long time series extraction ang change analysis of perennial and seasonal water surface in Heilongjiang Bain based on Google Earth Engine

Author
LIU Qing
Journal
Environmental Engineering
Class
Flood
Year
2021
Paper Keyword
Heilongjiang Basin, Google Earth Engine, Landsat, adaptive threshold, annual average water index
Abstract
Remote sensing image has wide spatial coverage and short update period, which is a feasible technology to extract water surface information in large area in time, and of great significance to the development, utilization, management and protection of water resources. In view of the influence of precipitation fluctuation within the year, this paper made full use of all available image data to construct the annual average water body index,so as to reduce the problem that was difficult to accurately reflect the water surface characteristics in a single period image. The Google Earth Engine ( GEE) remote sensing cloud platform was used to solve the problem of low efficiency of traditional image download and desktop processing of massive image data. Taking Heilongjiang Basin as the research area, taking Landsat Image as data source, combined with topographic data, annual water surface and seasonal water surface from 1987 to 2019 were extracted. The results showed that: 1) compared with the single period image data, he annual average water body index could reflect the time information of water surface more comprehensively, and the overall accuracy of water surface extraction was 95. 32%, in which the annual and seasonal water surface were 96. 59% and 94. 61% respectively; 2) compared with the existing data products, the annual water surface extracted in this paper was more continuous,complete and of better quality; 3) in recent 32 years,the annual water surface area of Heilongjiang Basin fluctuated greatly, showing a decreasing trend, with an average annual decrease of 14. 82 km2; the seasonal water surface was relatively stable, showing an increasing trend,with an average annual increase of 12. 81 km2.
    Sign in for comments!

Comment list ( 0 )