Do Water Quality Duration Curves and Flow Duration Curves Reflect the Socio-Hydrology Concept and Stochastic Behavior?

Código

I-EBHE0040

Autores

CAMILA DE CARVALHO ALMEIDA, Danieli Mara Ferreira, João Marcos Carvalho, Juliana Cristina Jansson Kissula, Marcelo Coelho, DANIEL HENRIQUE MARCO DETZEL, Cristovão Vicente Scapulatempo Fernandes

Tema

WG 1.03: Urban Water - Urbanization phenomenon & adequate water management

Resumo

Water quality management requires a combination of hydrological and water quality data analysis, which must be interpreted to propose actions for pollutant load removal. The regulation of water quality management in Brazil is based on concentration limits. Generally, the concentration of Biochemical Oxygen Demand (BOD) is adopted to represent the water quality variation and the estimation of BOD load to support the water quality planning, which includes proposing actions to improve the quality of the water. This paper aims to discuss how the BOD load to be removed can be estimated through duration curves (DC) loads calculated in distinct ways considering the reference of socio-hydrology concept and the need of discuss it in the context of participative process. The study case considers two monitoring points located in the Upper Iguassu Basin in the South of Brazil, an urban basin with a history of poor water quality conditions despite having considerably structured water resource governance. BOD and flow data monitored from 2005 to 2016 were utilized, reflecting a data series verified for RHIS (Randomness, Homogeneity, Independence, Stationarity) conditions. The concentration and flow monitored within the same day estimates the Observed Load (LODC). Considering that there are combined data (flow and BOD) for just 12 years, a complementary analysis with a longer flow series explores the representativeness of the short series used. Furthermore, there is a greater abundance of data concerning water flow than water quality, a transfer function between these two variables was analyzed as well as flow DC estimated by a hydrological model as complementary approaches. The Load Discharged duration curve (LDDC) was estimated using traditional per capita values for the numbers of inhabitants living in each monitoring basin contribution. The Load Supported (LSDC) based on the BOD concentration limits for each water quality class and the flow monitored despite the traditional that considers just the reference flow. The Load to be Removed (LRDC) was estimated based on the difference between LDDC and LSDC as well as between LODC and LSDC. The approach of LODC instead of year-to-year variation (commonly used) allows for visualizing the pattern of load variation over the twelve-year period, which can provide a more accurate insight into the load to be reduced to improve the water quality within the planning horizon, typically ranging from 10 to 30 years. The LRDC based on LODC and LDDC exhibit significant differences. According to LDDC, there is a higher load to be removed at the monitoring point located in the midst of the most urbanized area of the basin, whereas based on LODC, the loads to be removed are greater at the monitoring point representative of the basin outlet. Nevertheless, when observed from the 25% frequency level, both curves indicate a total to be removed close to each other (about 40 t BOD/day) for the highest quality class. Therefore, the DC approach can support decisions on the load to be removed based on the concept of return period associated with DC interpretation, which is already widely used for hydrological data.