EEWR Brown Bag Seminar with Bo Guo and Nathaniel Chaney, Graduate Students
Speaker: Bo Guo and Nathaniel Chaney, Graduate Students
Series: EEWR Brown Bag Seminars
Location: Engineering Quad E225
Date/Time: Friday, February 7, 2014, 12:00 p.m. - 1:30 p.m.
Title: A Vertically-Integrated Model with Subscale Vertical Dynamics for CO2 Storage
Speaker: Bo Guo
Conventional vertically-integrated models for CO2 storage usually adopt a vertical equilibrium (VE) assumption, which states that due to strong buoyancy, CO2 and brine segregate quickly, so that the fluids can be assumed to have essentially hydrostatic pressure distributions in the vertical direction. These vertical equilibrium models are accurate and computationally efficient as long as the VE assumption is satisfied. However, the VE assumption is inappropriate when the time scale of fluid segregation is not small relative to the simulation time, especially for the geological formations that have low permeability in the vertical direction, on the order of 10 millidarcy or lower (Court et al. 2012). By casting the vertically integrated equations into a multiscale framework, we develop a new vertically-integrated model that relaxes the assumption of vertical equilibrium, thereby allowing vertical dynamics to be modeled explicitly. The model maintains much of the computational efficiency of vertical integration while allowing a much wider range of problems to be modeled. Numerical tests of the new model, using injection scenarios with typical parameter sets, show excellent behavior of the new approach.
Title: Improving Spatial Heterogeneity in Macroscale Land Surface Models
Speaker: Nathaniel Chaney
Global land surface models have their roots in numerical weather prediction and global circulation models. As a result, they emphasize simulating land surface fluxes while oversimplifying hydrologic processes, limiting their usefulness for local scale applications. One of the main challenges is the adequate representation of landscape heterogeneity (e.g. soil properties, land cover, and topography). In this presentation, we will discuss how an increase in the reliability and accessibility of high-resolution continental scale land data sets can help solve this problem.
An enhanced statistical representation of land units (subsurface, surface, and vegetation) at the coarser grid scale can improve the representation of landscape heterogeneity and scalability between grid resolutions. To this end, adequate knowledge of the statistical distribution of soils, land cover, and topography within the coarser grid (including uncertainties) must be available from continental land datasets. Using available highresolution continental-scale land datasets (SSURGO, NLCD, and NED) over the Little River Experimental Watershed in Georgia, we will explore how the adequate statistical representation of fine-scale land units enables scalable model parameters and simulations using the TOPLATS hydrologic model.
Finally, we will discuss how these concepts are currently being implemented to couple the NOAH-MP land surface model to the Dynamic TOPMODEL hydrologic model. The goal is to provide a modeling framework that can maximize the use of high resolution land data sets, generalize to multiple climates and landscape, and provide a more formal link between semi-distributed and fully distributed land surface modeling.