TY  - JOUR
A1  - Soares Pereira, Francisco Jairo
A1  - Gomes Costa, Carlos Alexandre
A1  - Förster, Saskia
A1  - Brosinsky, Arlena
A1  - de Araujo, Jose Carlos
T1  - Estimation of suspended sediment concentration in an intermittent river using multi-temporal high-resolution satellite imagery
JF  - International journal of applied earth observation and geoinformation
N2  - There is a shortage of sediment-routing monitoring worldwide, despite its relevance to environmental processes. In drylands, where water resources are more vulnerable to the sediment dynamics, this flaw is even more harmful. In the semi-arid Caatinga biome in the North-east of Brazil, rivers are almost all intermittent and hydro-sedimentological monitoring is scarce. In the biome, water supply derives from thousands of surface reservoirs, whose water availability is liable to be reduced by siltation and sediment-related pollution. The goal of this research was to evaluate the potential of multi-temporal high-resolution satellite imagery (RapidEye) to assess the suspended sediment concentration (SSC) in the medium-sized intermittent Jaguaribe River, Brazil, during a 5-year period. We validated 15 one-, two- and three-band indices for SSC estimation based on RapidEye spectral bands deduced in the context of the present investigation and nine indices proposed in the literature for other optical sensors, by comparing them with in-situ concentration data. The in-situ SSC data ranged from 67 mg.L-1 to 230 mg.L-1. We concluded that RapidEye images can assess moderate SSC of intermittent rivers, even when their discharge is low. The RapidEye indices performed better than those from literature. The spectral band that best represented SSC was the near infrared, whose performance improved when associated with the green band. This conclusion agrees with literature findings for diverse sedimentological contexts. The three-band spectral indices performed worse than those with only one or two spectral bands, showing that the use of a third band did not enhance the model ability. Besides, we show that the hydrological characteristics of semi-arid intermittent rivers generate difficulties to monitor SSC using optical satellite remote sensing, such as time-concentrated sediment yield; and its association with recent rainfall events and, therefore, with cloudy sky.
KW  - Remote sensing
KW  - Sediment load
KW  - Dryland
KW  - Brazil
Y1  - 2019
U6  - https://doi.org/10.1016/j.jag.2019.02.009
SN  - 0303-2434
VL  - 79
SP  - 153
EP  - 161
PB  - Elsevier
CY  - Amsterdam
ER  -