5.12
Observing networks for ENSO prediction: Experiments with a simplified model
Rebecca E. Morss, NCAR, Boulder, CO; and D. S. Battisti
The Tropical Atmosphere-Ocean (TAO) array of moored buoys in the Pacific Ocean was deployed during Tropical Ocean-Global Atmosphere (TOGA) Program to provide real-time in situ observations for El Nino prediction. Since the early 1990's, the TAO array has regularly observed surface atmospheric winds, sea surface temperature, and subsurface ocean data at approximately 70 locations between 8N and 8S, 95W and 137E. Although the TAO array is a major source of data for understanding and predicting El Nino / Southern Oscillation (ENSO), little work has been performed to date on the best way to allocate observations for predicting ENSO. To address this issue, we have performed a series of observing system simulation experiments (OSSE's) with a simple dynamical ENSO model. The forecast model used is a linear variant of the Zebiak/Cane ENSO model, forced stochastically; the model is described in Thompson and Battisti (2000). The simulated observations are sea surface temperature (equivalent to surface winds in this model) and thermocline depth. Observations are taken regularly from the simulated "true" state and are incorporated into the simulated "model" state using a variational data assimilation system. We have compared forecast accuracy for a large number of simulated ENSO forecasts generated with a variety of observation densities, spacings, and distributions. To the extent that the simplified model dynamics represent ENSO dynamics accurately, the results begin to indicate how observations might best be allocated for ENSO prediction. The results also demonstrate that it is possible to evaluate the effectiveness of different observing networks for ENSO prediction within an OSSE framework.
Session 5, Testing and Simulation of Observing Systems: Part 1
Wednesday, 17 January 2001, 1:30 PM-4:45 PM
Previous paper