Innovative Smartphone-based Models for Estimating Water Quality Using RGB and XYZ Colour Spaces

Smartphones, although not empowered as water quality sensing devices, can be used to obtain spectral information about our water quality through their enhanced capability to capture images and process using applications (APPs) developed to work within given colour spaces. As the focus of this study, models were developed through two colour spaces used by existing smartphone APPs; RGB colour space by HydroColor and XYZ colour space by Citclops (now called EyeOnWater). For each colour space, a single band semi-analytical model and a logarithmic model were developed and evaluated to estimate turbidity and the concentration of suspended particulate materials (SPM) using smartphone images and validated with laboratory measurements. Although the two-colour spaces used in this research are different, they can be converted from one colour space to another. The models developed from the RGB colour space was suitable for a wider range of water bodies compared to the original models used by HydroColor APP. The XYZ colour space on conversion were also used to estimate turbidity and SPM compared to its original Citclops APP that estimated colour. Comparatively, the logarithmic models did well in the estimation of turbidity and SPM through the RGB and XYZ colour space with R2 ranging from 0.63 to 0.90 and RMSE from 5.86 to 20.18, to the semi-analytical model with R2 from 0.45 to 0.78 and RMSE from 0.83 to 27.25. In the broader context of water quality monitoring, these findings underscore the viability of smartphones in democratizing environmental data collection. The user-friendly XYZ colour space, requiring only water surface images, positions smartphone APPs as powerful tools for citizen engagement in water quality monitoring initiatives. This research contributes to advancing accessible and efficient methods for environmental monitoring, fostering a more citizen participatory approach towards safeguarding water resources, and addressing a novel approach to real-time water quality monitoring and mapping using widely accessible technology-smartphones.