RESUMO
Understanding and predicting the quality of inland waters are challenging, particularly in the context of intensifying climate extremes expected in the future. These challenges arise partly due to complex processes that regulate water quality, and arduous and expensive data collection that exacerbate the issue of data scarcity. Traditional process-based and statistical models often fall short in predicting water quality. In this Review, we posit that deep learning represents an underutilized yet promising approach that can unravel intricate structures and relationships in high-dimensional data. We demonstrate that deep learning methods can help address data scarcity by filling temporal and spatial gaps and aid in formulating and testing hypotheses via identifying influential drivers of water quality. This Review highlights the strengths and limitations of deep learning methods relative to traditional approaches, and underscores its potential as an emerging and indispensable approach in overcoming challenges and discovering new knowledge in water-quality sciences.
RESUMO
It is well known that groundwater arsenic (As) contamination affects million(s) of people throughout the Indus flood plain, Pakistan. In this study, groundwater (n = 96) and drilled borehole samples (n = 87 sediments of 12 boreholes) were collected to investigate geochemical proxy-indicators for As release into groundwater across floodplains of the Indus Basin. The mean dissolved (µg/L) and sedimentary As concentrations (mg/kg) showed significant association in all studied areas viz.; lower reaches of Indus flood plain area (71 and 12.7), upper flood plain areas (33.7 and 7.2), and Thal desert areas (5.3 and 4.7) and are indicative of Basin-scale geogenic As contamination. As contamination in aquifer sediments is dependent on various geochemical factors including particle size (3-4-fold higher As levels in fine clay particles than in fine-coarse sand), sediment types (3-fold higher As in Holocene sediments of floodplain areas vs Pleistocene/Quaternary sediments in the Thal desert) with varying proportion of Al-Fe-Mn oxides/hydroxides. The total organic carbon (TOC) of cored aquifer sediments yielded low TOC content (mean = 0.13 %), which indicates that organic carbon is not a major driver (with a few exceptions) of As mobilization in the Indus Basin. Alkaline pH, high dissolved sulfate and other water quality parameters indicate pH-induced As leaching and the dominance of oxidizing conditions in the aquifers of upper flood plain areas of Punjab, Pakistan while at the lower reaches of the Indus flood plain and alluvial pockets along the rivers with elevated flood-driven dissolved organic carbon (exhibiting high dissolved Mn and Fe and a wide range of redox conditions). Furthermore, we also identified that paired dissolved AsMn values (instead of AsFe) may serve as a geochemical marker of a range of redox conditions throughout Indus flood plains.