Estimation of Fluvial Suspended Sediment Concentration Using MODIS Sensor (Case Study: Hydrometry Station of Mollasani)

Document Type : Research Paper

Authors

1 Ph.D. student and Research Instructor of Soil Conservation and Watershed Management Research Institute

2 Head of GIS and RS Center, Faculty of Natural Resources, University of Sari, Iran

Abstract

     The fluvial suspended sediment concentration (SSC) which is directly and traditionally measured in the most times is costly and time consuming. Also, in some situations it has some human errors and also is calculated and recorded in a discontinuous time and spatial scale. MODIS data, because of high temporal resolution and accessibility are proper tools to estimate fluvial SSC.
In this research, the relationship between spectral reflectance of red and near infra red bands of MODIS sensor, their band ratios and their simultaneous fluvial SSC data have been evaluated.
Regarding with this, in a period of 9 years the SSC data of Mollasani hydrometry station (2003-2011) and simultaneous daily MODIS images (red and infra red bands with 250m spatial resolution) have been used.
 In order to understand the relationship between data, the correlation between the SSC and all simultaneous images for the entire period (all seasons together) was investigated and derived results showed a low correlation among them.In the next stage, the data of wet season (22 Nov to 20 Apr) for entire statistical period were evaluated and results showed a high correlation between SSC and water spectral reflectance
 (r= %88).To make a regression relation, some linear and nonlinear relations were statistically considered and eventually a linear relation with having suitable parameters (R2= 0.77, Sig.= 0, F= 43) was calibrated
 (a= 1.27, b= 0.242, n= 15). The calibrated relation then was validated using leave one out cross validation (LOOCV) method (RMSE= 55 mg/l, RRMSE= %26). In summary, despite the insufficient number of fluvial SSC data, the existence of cloud and aerosol in the images and also other influencing parameters in spectral reflectance of Karun river, the results showed good performance of MODIS images in estimating daily fluvial SSC in large rivers in wet season.
 
 

Keywords