Open AccessRegional climates in GCMs. Final report
Author(s) -
Robert G. Crane
Publication year - 1995
Language(s) - English
Resource type - Reports
DOI - 10.2172/491353
Subject(s) - climatology , snowpack , precipitation , snow , general circulation model , environmental science , climate model , scale (ratio) , circulation (fluid dynamics) , structural basin , atmospheric circulation , meteorology , climate change , geology , geography , cartography , paleontology , oceanography , physics , thermodynamics
This research describes empirical methods developed to obtain short-term, regional results from global climate models. Observational data sets were compared to the GENESIS climate model; spatial and temporal variability were examined to validate the circulation model on the synoptic scale. A feed-forward neural network was used to determine transfer functions for circulation-precipitation relationships. The empirical methodologies derived were then applied to the analysis of mountain snowpack in the upper Colorado Basin. The comparison of observational data and the model showed that the synoptic scale circulation of the GENESIS model is realistic over the eastern United States; however, the model features are displaced south by about five degrees and actual pressures in the model are much lower than observed pressures. Preliminary results from the neural network produced correlations between observed and predicted rainfall of about 0.7 to 0.8, depending on the net configuration. Similar results were obtained for the upper Colorado Basin study in the prediction of winter snowfall. 6 refs., 3 figs