Revista de Teledetección (Dec 2017)
Reflectances of SPOT multispectral images associated with the turbidity of the Upper Gulf of California
Abstract
The use of satellite images for the observation and measurement of marine turbidity has been developed mainly with ocean colour sensors, such as MODIS. These images have a maximum spatial resolution of 250 m in their visible and infrared bands. In this research, images of the SPOT sensors were chosen as an alternative to overcome this limited spatial resolution. The objective was to prove the suitability of SPOT to measure turbidity in areas with great spatial variability. As a first step, all the images were standardized and the SPOT wavelength that had the largest association in the Principal Component Analysis was chosen (PCA). The results show that the bands of a SPOT multispectral image are highly redundant. The wavelength of the 610-680 nm (S2610-680) obtained the best association in 89% of the 73 images analysed. The SPOT reflectance (Rrs) (S2610-680) was compared with MODIS 620-670 nm (M1620-670), which has already been tested in other research and has proved to be adequate for measuring turbidity. Both sensors performance was similar for low and moderate reflectance but for high reflectance, SPOT (S2610-680) had a better performance than MODIS (M1620-670). Additionally, the SPOT Rrs (S2610-680) was associated with standardized Secchi disk depth data, which were measured in situ, to check SPOT suitability. SPOT Rrs (S2610-680)images were classified into: 1) cold or warm season, 2) spring tide or neap tide and 3) water flux or reflux. These constructed scenarios allowed to see the result of the Standardized Space Anomalies, which showed the continuous presence of low and medium values in the most oceanic region of the Upper Gulf of California (UGC) and very high values in all the scenarios in the intertidal zone. This research has shown that SPOT Rrs (S2610-680) is useful for observing, differentiating and measuring turbidity patterns in areas with very high spatial variability.
Keywords