Errors Analysis Of Spectral Measurement In High Turbidity Estuarine Waters

The in-situ spectroradiometry measurement data is the basis for establishing a high-precision spectral inversion model between the satellite sensors and land covers.However,because of the unsteady geometry of shipborne above water and the dynamic factors,such as flow rate and suspended sediment content,there is a large error in spectroradiometry measurement.Therefore,considering the influence of environment factors and instrument settings on the spectroradiometry measurement above the high-turbidity water surface,it is of great significance to explore the improvement methods and approaches to improve the accuracy of inversion accuracy and to understand the optical properties of the case II water.In the study,we compared the spectral data collected from two spectroradiometers in the field,analyzed the differences of the data from the above-water radiation measured by spectroradiometers at different integration times under the influence of flow velocity,and combined with the suspended sediment concentration and flow velocity measured in situ,the result shown as follows:(1)The spectral values by the Hypersas measured overall higher than the ASD measured at the integration time of 34 ms,the difference between the two minimum,average relative differences were 22.70%,18.73%,15.01%,14.07%,13.05%,11.93%,9.58% and 9.85% in the typical band 412 nm,443 nm,490 nm,510 nm,531 nm,555 nm,670 nm and 750 nm respectively.Based on the correlation between the spectral values measured by Hypersas and ASD and the suspended sediment,it is suggested that using the ASD spectroradiometer for field measurements.(2)The flow rate has an influence on the field spectroradiometry measurement.When the flow rate exceeds 1 m/s,the larger measurement deviation comes out when the current velocity becoming more higher.The quality of the data was verified byusing surface suspended sediment and remote sensing reflectance regression fitting accuracy.When the current velocity changes in wide range,it is recommended to use the integral time 34 ms on the measurement.(3)The spectrum has a higher correlation with the suspended sediment concentration in the water with high suspended sediment concentration value,while the suspended sediment concentration is low,the correlation between these two is low.Comparing the remote sensing inversion models of two waters,it is found that the best modeling band and modeling parameters are not the same.Therefore,it is not recommended to use the same model to invert the waters with different suspended sediment condition.(4)Based on the above studies,we select the integral time of 34 ms to model the data of North Channel and South Passage.Using the satellite images with different remote sensors to invert the suspended sediment concentration in the Yangtze Estuary,the results obviously shown the variation and distribution of suspended sediment concentration in the Yangtze Estuary with the tides and seasons changing.