Midwest Mathematics and Climate Conference - Day 1 Afternoon Session

View Link (HTM)

Licensed according to this deed.

Published on

Abstract

  • Graham Feingold, National Oceanic & Atmospheric Administraton
    Dynamical System Analogues to Cloud Systems

    • Shallow convection exhibits fascinating cellular structures at scales of a few to several hundred kilometers. Configurations of relatively cloud free open cell states, or much cloudier closed cell states can be clearly distinguished. The difference in cloud fraction between these states has important consequences for the amount of shortwave energy reflected back to space, and therefore the role of these clouds in modifying energy flows in the climate system. Transitions between these states are a topic of active study. Studies to date have identified the fundamental processes responsible for these states and transitions between them. Here we take a different view of the problem and apply dynamical systems analogues to either the entire cloud system, or to individual (coupled) clouds within the system. We show that simple sets of equations mimic some of the key properties of the solutions derived from computationally intensive large eddy simulation.

  • Sean Crowell, University of Oklahoma
    The Carbon Cycle as a Leading Order Uncertainty in Climate Prediction​

    • The uncertainty in future projections of the earth’s climate due to representations of the water cycle (e.g. clouds) and aerosol processes is well known. A widespread effort under the auspices of the Department of Energy’s Atmospheric Systems Research (ASR) program seeks to provide observational constraints on process models in order to improve this misrepresentation in Earth Systems Models (ESMs). In this presentation, we will introduce the basic concepts of terrestrial ecology models, and demonstrate that the uncertainties present in these models yield climate prediction uncertainties of the same order as those coming from cloud and aerosol processes. We will give examples of efforts currently underway to use data assimilation techniques to constrain processes in these models, and describe proposed attempts to utilize an ever growing data set to improve models.

  • Vincent Larson, University of Wisconsin, Milwaukee
    Monte Carlo Modeling of Clouds and Precipitation

    • Weather forecast and climate models cannot explicitly resolve either small clouds or precipitation processes. However, their effects can be estimated. In particular, we will discuss a method in which the probability density function (PDF) of cloud fields is estimated, the PDF is sampled by Monte Carlo methods, and those samples are used to estimate precipitation process rates. Challenges for the method include modeling the PDFs accurately and reducing sample noise.

  • Nathaniel Brunsell, University of Kansas
    Disentangling Climate and Land-use Impacts on Carbon and Water Fluxes

    • Regional climate and land cover interact in a complex, non-linear manner to alter the local cycling of mass and energy. It is often difficult to isolate the role of either mechanism on the resultant fluxes. Here, we attempt to isolate these mechanisms through the use of network of 4 Ameriflux eddy covariance towers installed over different land cover and land use classes along a pronounced rainfall gradient. We investigated the impact of rainfall variability, drought, and heat waves on the water and carbon budgets using data analysis, remote sensing, and modeling approaches. Through the use of a wavelet multiscale decomposition and information theory metrics, we have isolated the role of environmental factors (temperature, humidity, soil moisture, etc.) on the fluxes across the different sites. By applying a similar analysis to model output, we can assess the ability of land-surface models to recreate the observed sensitivity. Results indicate the utility of a network of measurement systems used in conjunction with land surface modeling and time series analysis to assess differential impacts to similar regional scale climate forcings. Implications for the role of land cover class in regional and global scale modeling systems will also be discussed.
  • Simon de Szoeke, Oregon State University
    Radiative Effects and Feedbacks of Marine Stratiform Clouds

    • By reflecting sunlight, low clouds cool the climate. Because they increase for cooler SST, reflecting more sunlight, further cooling SST; marine low clouds have a positive climate feedback. This low cloud feedback is represented with such variety in climate models that the fate of low clouds is a leading source of uncertainty for climate projections. A simplified model for SST and air temperature interacting with marine atmospheric boundary layer (ABL) clouds allows exploration of temperature and cloud equilibria and stability regimes. Sunlight warms the ocean mixed layer; turbulent surface fluxes (and longwave emission) cool it. The turbulent fluxes depend on the mixed layer temperature, so there is an equilibrium temperature at which the fluxes balance the solar warming. Linearizing the temperature dependence of the turbulent (and longwave radiative) surface fluxes about the equilibrium temperature gives an adjustment time scale of 60 days for the turbulent fluxes to relax the ocean mixed layer to the equilibrium temperature. Surface and entrainment fluxes relax the ABL temperature and humidity, on a time scale of 2 days, towards an equilibrium value between the saturated surface and the relatively warm dry free troposphere. The positive radiative feedback of ABL clouds opposes the damping effect of the turbulent fluxes, increasing the damping time scale of the ocean mixed layer to 150 days. For reasonable parameter choices, the positive cloud feedback to the ABL temperature overwhelms the damping due to the turbulent fluxes, resulting in local instability for intermediate cloud amounts of 40-60%, where clouds are most sensitive to changes in the mean relative humidity. For either clearer or more overcast conditions than this, the turbulent fluxes stabilize the system. The instability at intermediate cloud amounts suggests a qualitative behavior for the marine ABL to be either nearly clear or nearly overcast. Implications will be discussed for observed boundary layer clouds, which are forced by seasonal, synoptic, and diurnal variability.

  • Mark Taylor, Sandia National Laboratories
    Scalable Transport Algorithms for Global Atmospheric Models

    • I will give an overview of several tracer transport algorithms being developed for the HOMME spectral element dynamical core used in the Community Atmosphere Model (CAM). The algorithms include Eulerian, semi-Lagrangian and hybrid methods, with a focus on algorithms that can run near the limits of strong scaling on parallel computers. They span a wide range of tradeoffs between local computational density, parallel communication latency and bandwidth. A key numerical issue is how to obtaining tracer mass consistency with respect to the air density from the continuity equation. The continuity equation is coupled to the momentum equation and in the case of HOMME, is solved with spectral/finite element methods. For HOMME, we address this issue by advecting the tracers with a mean mass flux and show how to extract this mean flux from the spectral element method.

Submitter

Colin James Grudzien

University of North Carolina at Chapel Hill Mathematics

Tags