H43B-1489
Modeling Soil Water in the Caatinga Tropical Dry Forest of Northeastern Brazil

Thursday, 17 December 2015
Poster Hall (Moscone South)
Cynthia Wright1, Brad Wilcox1, Eduardo S. de Souza2, Jose Romualdo de S. Lima3 and Jason West1, (1)Texas A & M University College Station, College Station, TX, United States, (2)Universidad Federal Rural de Pernambuco, Serra Talhada, Brazil, (3)Universidad Federal Rural de Pernambuco, Garanhuns, Brazil
Abstract:
The Caatinga is a tropical dry forest unique to northeastern Brazil. It has a relatively high degree of endism and supports a population of about 20 million subsistence farmers. However, it is poorly understood, under-researched and often over-looked in regards to other Brazilian ecosystems. It is a highly perturbed system that suffers from deforestation, land use change, and may be threatened by climate change. How these perturbations affect hydrology is unknown, but may have implications for biodiversity and ecosystem services and resiliency. Therefore, understanding key hydrological processes is critical, particularly as related to deforestation. In this study, Hydrus 1D, which is based on van Genuchten parameters to describe the soil water curve and Richard’s Equation to describe flow in the vadose zone, was used to model soil moisture in the Caatinga ecosystem. The aim was 1) to compare hydraulic characterization between a forested Caatinga site and a deforested pasture site, 2) to analyze inter-annual variability, and 3) to compare with observed soil moisture data. Hydraulic characterization included hydraulic conductivity, infiltration, water content and pressure head trends. Van Genuchten parameters were derived using the Beerkan method, which is based on soil texture, particle distribution, as well as in-situ small-scale infiltration experiments. Observational data included soil moisture and precipitation logged every half-hour from September 2013 to April 2014 to include the dry season and rainy season. It is expected that the forested Caatinga site will have a higher hydraulic conductivity as well as retain higher soil moisture values. These differences may be amplified during the dry season, as water resources become scarce. Deviations between modeled data and observed data will allow for further hypothesis to be proposed, especially those related to soil water repellency. Hence, these results may indicate difference in soil water dynamics between a forested and non-forested site which will have implications for land use and management strategies that promote water resource conservation and availability.