Lake water volume calculation with time series remote-sensing images

Abstract

The volume of water in lakes is commonly estimated by combining data of water level variations with accurate bathymetry and shore topographic maps. However, bathymetry and shore topography data are often difficult to acquire, due to high costs for labour and equipment. This article presents an innovative method for calculating lake water volumes by using long-term time series remote-sensing data. Multi-spectral satellite remote-sensing images were used to map a lake’s water surface area. The lake water surface boundaries for each year were combined with field-observed water levels to generate a description of the underwater terrain. The lake water volume was then calculated from the water surface area and the underwater terrain data using a constructed TIN (triangulated irregular network) volume model. Lake Baiyangdian, the largest shallow freshwater lake in the North China Plain, was chosen as the case study area. For the last 40 years the water levels of Lake Baiyangdian have reflected multiple dry and wet periods, which provide a good data series for the study of the proposed method. Archived Landsat MSS/TM/ETM+ and HJ-1A/B images from 1973 to 2011 were used as the basic data. The NDWI (normalized difference water index) and MNDWI (modified NDWI) were used to map the water surface of the lake, and the lake water volumes were calculated with the 3D Analyst tool of ArcMap 9.3. The results show that the estimated water volumes from remote-sensing images were very consistent with the volumes derived from the fitted equation of the lake storage capacity curve based on observed data.

Acknowledgements

This study was funded by the National Natural Science Foundation of China (No. 41101401 and No. 91025007) and the Non-profit Research Programme of the Ministry of Water Resource (No. 201101015). We are grateful to the US Geological Survey Earth Resources Observation and Science Center (EROS), the Centre for Earth Observation and Digital Earth, Chinese Academy of Sciences, and the China Centre for Resources Satellite Data and Application for providing the Landsat MSS/TM/ETM+ and HJ-1A/B images. We are profoundly grateful to Professor Ian Willett from the University of Melbourne for carefully reviewing the article and providing valuable advice.