| Event Description: Paul Krause, Climate and Global Dynamics Laboratory, University Corporation for Atmospheric Research (UCAR) Influence sampling, strong predictability and analysis of non-observable trailing variables A novel sampling scheme for the state of an evolutive system of ordinary differential equations (o.d.e.), the Influence Sampling (IS) scheme, is presented for systems with local determining variables. The concept of local determining variables is an extension of the concept of determining variables on the attractor. The IS is a discrete control scheme for trailing variables of an evolutive system of o.d.e.. It adapts sample values of the trailing variables to discrete control values of the local determining. In predictability tests conducted on solutions of the L63 (Lorenz, 1963), the IS scheme is found to increase the period of strong predictability of the L63 system and the accuracy and confidence of the estimate in this period. In recovery tests conducted on solutions of the moist-convective rotating shallow water (mcRSW) system of equations and a 3D climate model, the IS scheme is found to recover the mcRSW's layer thickness and humidity and the climate model's convection and humidity with great accuracy when the mcRSW wind is provided every 12 hours and the climate wind, temperature and density |
| Location Information: Main Campus - Engineering Office Tower (View Map) 1111 Engineering DR Boulder, CO Room: 226: Applied Math Conference Room |
| Contact Information: Name: Ian Cunningham Phone: 303-492-4668 Email: amassist@colorado.edu |