Title of article :
River discharge monitoring using satellite missions: Sentinel-1, Sentinel-2, and Sentinel-3 (Case study: The Karun River, Iran)
Author/Authors :
Tayfehrostami, Arash School of Surveying and Geospatial Engineering - College of Engineering - University of Tehran, Tehran, Iran. , A. Ardalan, Alireza School of Surveying and Geospatial Engineering - College of Engineering - University of Tehran, Tehran, Iran. , Pourmina, Amir Hossein School of Surveying and Geospatial Engineering - College of Engineering - University of Tehran, Tehran, Iran.
Abstract :
With a length of 950 km, Karun River is the longest river in Iran. In this study, we aimed at application of
Sentinel-3B satellite altimetry data as well as Sentinel-1 and Sentinel-2 satellite imagery for the estimation
of Karun River discharge and validation with the in-situ data. Knowing that Level-2 altimetry data are not
reliable for rivers and shallow waters, we opted to re-track the waveforms of Level-1B Sentinel-3B
mission data and to test several re-tracking techniques for this purpose. The results showed that the
threshold algorithm, with threshold of 90%, improves the accuracy of the time series of water level by
7.05% and increases the correlation with the in-situ gauge data by 12.7% as compared with those obtained
via Level-2 data based on OCOG that was identified as the optimum re-tracker in this case. Next, from the
estimated time series of the river’s water level, the time series of Karun River discharge were evaluated in
order to constitute our discharge estimation based on Sentinel-3B satellite altimetry data, which further to
be compared with the discharge that we calculated using satellite imagery of Sentinel-1 and Sentinel-2,
while taking the in-situ data as the benchmark. The river’s discharge time series obtained from the
altimetry data resulted in RMSE value of 852.31 m3
/s, NSE coefficient of 0.19 m3
/s, and correlation of
62.40% with the in-situ river discharge time series. On the other hand, the river discharge time series
obtained from satellite imagery of Sentinel-1 mission resulted in RMSE value of 165.06 m3
/s, NSE
coefficient of 0.94 m3
/s, and correlation of 97.12%, and Sentinel-2 mission the RMSE value 264.23 m3
/s,
NSE coefficient of 0.81 m3
/s, and the correlation of 97.32% with in-situ data. The overall results of this
study indicates that various Copernicus satellites missions have good potentiality for Karun River
discharge monitoring.
Keywords :
Karun River , Retracking , Returned waveforms , Satellite radar altimetry , Satellite Imagery
Journal title :
Earth Observation and Geomatics Engineering