Stepwise aggregation method for the WQI-CCME to generate a water quality profile for a long river: case study-São...

This paper presents a new aggregation method to calculate the WQI-CCME, named stepwise aggregation method, to communicate the water quality of the São Francisco River, Brazil, a river of national expression, with more than 2700 km long. Differently from the original proposal for the WQI-CCME, the new aggregation method uses data from a single monitoring campaign, to avoid the loss of data, as many variables are not systematically measured in all monitoring campaigns. The stepwise aggregation method also requires to combine data from a minimum of four monitoring locations per section of the river, to attend the statistical requirement, a proposal already presented in the literature. After using the first four points, the next calculation requires to abandon the first point (upstream) and to include the next point (downstream), and the sequential calculation will generate the results to build the river water quality profile. In the present work, with 31 monitoring points, the stepwise aggregation method allowed to calculate the WQI-CCME for 25 sections. The stepwise method has the advantage of multiplying the number of index calculations, allowing to construct a more uniform water quality profile for a long river. The choice of using the WQI-CCME in this work, is because this index has many successful applications all over the world, favored by its flexibility of parameter selection, ease of calculation, adaptability to legal requirements of different locations and different water uses. The WQI-CCME profiles were compared with those for the WQI-CETESB, an index derived from the WQI-NSF and extensively used by the Brazilian Monitoring Programs. The comparison demonstrated similarity, only if the same number of parameters were used, but WQI-CETESB gives better water quality, similarly reported for the WQI-NSF, which systematically classify the water quality in the superior range. In this paper, the calculation of the WQI-CCME, with 38 and 20 parameters, confirmed the dependence of the WQI-CCME factors (F1, F2) on the number of variables considered. The calculation with more variables (38) gave better results for the water quality than the calculation with less variable (20). Thus, while the application of the WQI-CCME has a variety of advantages, it requires careful evaluation about the variables to use, and the objectives for the index calculation, as considered by many authors.