Modeling versus remote sensing techniques for water quality monitoring in deltaic coastal lake in Egypt

Abstract

Coastal zone of Nile Delta experience challenges with excessive developments and concerns with climate change impact. Ensuring sustainability and strengthened adaptive capacity require an integrative base of information as well as regular monitoring for management planning. Difficulties with spatial coverage and frequency of field-based data collection triggered the need to seek other methods to fill the gaps. While employing numerical modeling proved acceptable performance in certain water quality predictions, the advancing Remote Sensing (RS) techniques provide a vast pool of spatial, temporal and multispectral data that are useful in information retrieval. This research investigates reliability degree of remotely sensed (Landsat-8) water quality parameters versus Model (Delft3D) predictions. Through an application in a deltaic coastal lake, the study presents a comparative analysis that address aspects such as pre-processing requirements, provided inputs, constraints in coverage, processing complexity as well as accuracy of concluded results for modeling versus space-based remote sensing techniques. Despite the dynamic nature and zonation featuring the water body under consideration, accuracy of model-predicted and satellite-retrieved water quality parameters [Total Suspended Sediments (TSS), Chlorophyll-a, Ortho-Phosphate (PO4), Biological Oxygen Demand (BOD), Dissolved Oxygen (DO), Temperature], proved acceptably correlated (R2 ranges0.73-0.86 and 0.69-0.87); respectively. However, the comparative analysis highlighted positive and negative issues of each technique and, accordingly, established guidance factors to consider by decision makers in prioritization and selection criteria based on management requirements. With proved effectiveness, regular assessment and wide spatio-temporal water quality database is feasible with complementary functioning of modeling and remote sensing techniques.