It depends on the environment and activities around the areas of interests. The variables to be selected in the modeling should be significant and relevant to your research objectives.
There are few studies that attempt to model both flow and water quality taking into account the whole complexity of the physical, chemical, and biological processes involved. Moreover, urban water quality studies need to combine hydrological modeling of natural surfaces with the performance of urban man-made structures and impervious areas in a comprehensive hydrological modeling approach.
Pollution due to industrial effluents and domestic sewage
Mining
Recreational activities
etc.
Water is polluted by many factors among which industrial wastes are the most important. Beside industrial wastes other factors include herbicides, pesticides and atmospheric pollutants. Pathogen in polluted water causes serious diseases in humans. The whole ecosystem of water bodies is disturbing due to water pollution.
To treat industrial wastes there should be special industrial waste treatment plants with every industry. Similarly there should also be urban runoff to remove pollutants from runoff and to prevent floods. Toxic pesticides and Herbicides should be replaced with nontoxic ones or Pesticides should be replaced with biological control.
To build a city, we try to turn it into a desert by getting all of the rain to runoff, and all of the runoff to go away quickly. This introduces a range of sediments, dissolved chemistry, and thermal impacts into the receiving stream. Modelling for all of these simultaneously is complex - even 'each' of these is fairly difficult. Collecting baseline data, collecting calibration data, collecting rainfall intensity and durations, collecting baseflow data - all of these require different methods, and make different assumptions.
To collect baseline data, we once installed a sensor under a bridge near the outlet of the stream. It was vandalized about a dozen times in 4 months - eventually we put up a remote camera to see who was doing it, and it turned out to be beavers pushing it around. People will also disturb sensors, so you end up putting data loggers in hard to reach or well protected locations, which may make the readings less precise.
Once we were collecting information immediately downstream of an armed forces base in an urban setting. We were instructed to collect information on only the precise chemistry that they had given permission for - they didn't want us to find anything nasty that they were playing with there. We ended up collecting information that was of significantly less value, because there were stake holders that didn't want a full awareness of water quality.
We were doing a geomorphic assessment of a clayey site, and there was a locallized collapse of the clay bank, in a very non-typical manner. It turned out that over the years, salt had been dumped along the edge of a parking lot, and it had reacted in some way with the clay soils, significantly reducing it's strength. There can be very complex interactions between the soils and water quality, and the apparent problem (a clay sediment source) might be quite different from the actual problem (heavy salt loading, weakening the clay).
Another time, I was on the review-agency side of things, and we were brought in to help a developer understand what was happening at their site. Clay soils, but sand and fine gravel was accumulating in the channel, causing the watercourse to rise and widen, increasing it's meander amplitude and rate. Walking upstream, we found a city pathway that was built deep in the flood plain, and washed out every year. And the city would maintain it by adding several truck loads of stone dust and fine gravel after every spring freshet. No strength, but cheap - and it filled the channel, 2km downstream. Monitoring and modelling don't tell you where the problem really is, only that when there are differences between what you expect and what you are actually seeing, there is likely a human factor that is unaccounted for.
I'll leave it there, for others to provide their experiences.
In an attempt to produce a reliable source of water, the water quality model QUAL2Kw was put to through calibration and validation tests. This model was implemented in the river Tungabhadra of India. This worked quite well for the project at hand but showed certain irregularities. Different Quality Management methods exist currently and for this project, their impact in DO concentrations were examined based on: 1) The ability to augment the flow; 2) The ability to oxygenate the source and; 3) The ability to modify pollution loads. With the impacts realized, the study helped us under-stand that local oxygenation is effective when considering the increment of Do levels. There was also the understanding that a combination of the fac-tors (flow augmentation, pollution load modification and source oxygenation) is a necessity as far as the minimum DO concentrations must be realized. With this quality model results, there is confidence that the implementation of the QUAL2Kw model will be an appropriate choice for future river water quality policies.
Article Utilization of Water Quality Modeling and Dissolved Oxygen C...