You can use landuse change data for the past to simulate runoff and compare with runoff and river flow patterns........................ Urban development and Landuse will change Tc value ..................... 

use hspf 

Buying a water quality sonde such as from Global Water may help you, and with a boat or raft might provide some mobility potential for sampling various sites.  Costs for much of this type work can be substantial. Otherwise it is water sampling and flow measuring to determine water quality changes with discharge rate.  Apply standard methods for analysis of water and wastewater (reference book).  Various analysis techniques exist for analyzing most metals and I believe quite a few could be addressed with atomic absorption or flame emission spectrometry in the lab, lots of samples, preservation, diluting, etc to get results.  Automatic samplers of some help to catch storm flushes.  Large rivers are difficult to sample and be sure you are in a well mixed section well below any pollutant inputs for proper mixing.  You might be able to compare water quality above urban area and water quality below, or just sample tributaries from urban runoff if no direct discharges. If you are more interested in organism effects you might use biological indicators such as above pollution, within pollution areas and below pollution areas.  Aquatic insects, fish assemblage and some algae are rated for their pollutant tolerance, and perhaps other indicators such as abrupt species changes.  If you are not familiar with aquatic insect surveys, you might look at our report on using them for insecticide indicator for spruce budworm project (1984 report with moth on the cover).  Again, river size will depend on what you might consider.