Causality and persistence in the water availability dynamics

• Grant number: 18/19222-7
• Support type: Scholarships in Brazil - Master
• Effective date (Start): December 01, 2018
• Effective date (End): February 29, 2020
• Field of knowledge: Physical Sciences and Mathematics - Geosciences
• Principal Investigator: Edson Cezar Wendland
• Grantee: Kalyl Gomes Calixto
• Home Institution: Escola de Engenharia de São Carlos (EESC). Universidade de São Paulo (USP). São Carlos, SP, Brazil
• Associated research grant: 15/03806-1 - Water availability and quality threats in a Guarani Aquifer System outcrop zone, AP.BIOEN.TEM

Abstract

A better understanding of hydrological processes at basin scale is a challenge to achieve a more sustainable water resources management, especially towards predicting different behaviours under climate or land use change scenarios. Despite the evidence that some hydrological variables are resulting of long memory processes, just a few works explore this pattern along with the modelling process. Thereby, this proposal seeks to build a conceptual hydrological model to represent the water availability dynamics of an experimental study basin, taking advantage of prior characterization of the causal relationships between hydrometeorological time series. A hypothesis to be tested is that relevant information about water storages in the basin may be revealed by combining alternative interpretation tools, such those based on multiresolution analysis and on nonlinear causality tests. The study area, Onça Creek basin (central region of Sao Paulo state), is an upland watershed, located on an outcrop area of the Guarani Aquifer System, under predominant agricultural use, which has been hydrologically monitored since 2004. Based on locally registered data, the precipitation and evapotranspiration processes shall be quantified and spatially interpolated using different temporal scales. The streamflow components, baseflow and direct runoff, shall be estimated by using a recursive digital filter. A set of time series composed by total streamflow, baseflow, precipitation, evapotranspiration, and groundwater level of representative watershed wells will be integrally analysed by using a multiresolution approach, oriented to identify cause-effect relationships, and also to quantify the persistence between processes. The model-building will be carried out in a flexible framework that allows for behavioural adaptation to represent the uniqueness of hydrological systems, and to reproduce its different signatures. Among the expected contributions we highlight: a better hydrological understanding of the study area; and insights about how to build an effective water availability forecast tool. (AU)