A regional TSS retrieval model incorporating diffuse attenuation coefficient for a macro-tidal environment using satellite images: Gulf of Khambhat, India

Abstract

A study of water quality parameters in the highly dynamic macrotidal region (ranging from 3 to 12 m) is required to reduce the high cost and labour intensity of field surveys. With advances in remote sensing technologies, it has become possible to utilise the enhanced spatial as well as temporal resolution of satellite image observations to estimate water quality parameters more accurately. This study proposes to develop a new regional model to estimate Total Suspended Sediment (TSS) by integrating Moderate Resolution Imaging Spectroradiometer (MODIS) images and field measurements for the dynamic region of the Gulf of Khambhat, India. This study uses regression analysis to develop and calibrate a new model using remote sensing reflectance and the diffuse light attenuation coefficient from MODIS images, along with field-measured TSS (obtained in December 2012 and March 2013). March 2015 data was used to validate the model. The regression coefficient (R2) of 0.8168, the root mean square error (RMSE) of 3.1 mg/L, and the mean relative error (MRE) of 0.3815 % are characteristics of the model, which will enable estimation of a wide range (60 mg/L to 2000 mg/L) of TSS values. The unique feature of the model is that results are obtained at 500 m resolution, using diffuse attenuation coefficient values calculated from MODIS images at 500 m instead of 1000 m resolution. With the availability of higher-resolution satellite images, the derived equation has been used to estimate TSS for recent years, particularly during the COVID-19 lockdown, when field data collection was not possible. Sentinel 2 MultiSpectral Instrument images have been used for the Gulf of Khambhat, and the derived TSS equation was applied to the study before, during, and after the COVID-19 lockdown.