Spatio-temporal, geospatial, and time series analysis of water quality estimation using Landsat 8,9, Sentinel-2, and MODIS series for the region of India: A Google Earth Engine based web-application

Abstract

<p>Due to rising water quality-related issues, a periodic and continuous monitoring system is mandatory for inland water bodies. Water quality estimation is essential for water resource management and the sustainability of riverine ecosystems. Existing in-situ, field-based, and wet laboratory estimations, although precise and accurate, account for the lack of spatial and temporal variability and represent point sampled assessment. With a high temporal resolution and fine spatial resolved scaling, remote sensing data, including the Landsat-8,9 series, and Sentinel-2 series, consecutively provide high-spatio-temporal resolution observations for real-time analysis. The Google server and cloud-based Google Earth Engine (GEE) platform support image collections, atmospherically-radiometrically corrected imagery, and large-data processing. Taking the inland waterbodies of Delhi as the study area, this study is carried out in GEE to (i) design, inquire and pre-process all Landsat and Sentinel series observations that coincide with in situ measurements; (ii) extract the spectra to develop empirical models for water quality parameters and (iii) visualize the results graphically using geospatial distribution maps, time-series charts, and create a web-application. Water quality parametric analyses were conducted for Optically Active constituents (OAC), i.e., chlorophyll-a, suspended solids, and turbidity. Validation with an independent site location is the next area of study for estimating the predicted and observed values. Spectral characteristics show correlation and similarity with the field data and active optical constituents. Besides visualizing long-term spatial and temporal variabilities through thematic maps and time-series charts, anomalies such as eutrophication at specific sites can also identify using the models developed. An online application is in progress to allow users to explore and analyze water quality trends using the latest Landsat-9 dataset. Integrating remotely-sensed images, in situ measurements, and cloud computing can offer new opportunities to implement effective monitoring programs and understand water quality dynamics.</p>