The Application of High Resolution Dynamical-Numerical Models as a Tool to Infer Climate Statistics: January Simulations.
Abstract
Overall, the model was able to reproduce the long term means and variability. A slight bias in pressure has lead to a re-evaluation of the initialization fields to ensure proper mass balances (which should remove the bias). Also the semi diurnal tide in pressure was missing but further research found that this effect could be included by a post-processing procedure. The 40 km temperature and dew point fields generally showed diurnal trends closer to observed. This is likely due to the "assimilation" of observed temperatures into the 40 km simulations but not the 10 km simulations. As expected, variability is generally more accurately reproduced in the 10 km simulations. However, the fields dominated by large scale variability (i.e., pressure arid temperature) showed on minimal improvements. The most notable improvements in capturing the observed variability were associated with the wind field. Often the 40 km simulations only capture a small portion of the total variability of the wind field (typically 30%), whereas the 10 km simulations captured nearly half of the observed variability. Within the model resolution the shape of the frequency distributions was generally well represented by the model. However some discrepancies in temperature and dew point temperature has lead to further investigations of the model's moist thermodynamic processes.
Document Details
- Document Type
- Technical Report
- Publication Date
- May 28, 1998
- Accession Number
- ADA369473
Entities
People
- Charles E. Graves
- John Zack
Organizations
- Saint Louis University