Font Size: a A A

Sensitivity of seasonal forecast to the parameterization of land surface processes

Posted on:2002-02-24Degree:Ph.DType:Dissertation
University:The Florida State UniversityCandidate:Stefanova, Lydia BFull Text:PDF
GTID:1460390014950779Subject:Physics
Abstract/Summary:
A detailed land surface parameterization scheme, SiB2, using satellite derived vegetation cover greenness, is incorporated into the Florida State University global atmosphere circulation model, replacing the original simple land surface treatment. The effect of the new land surface scheme on the model performance is assessed in terms of changes in model climate, changes in seasonal forecast skill and changes in predictability of summer time seasonal forecast. Three sets of three-member ensemble forecasts were made for the 1987 and 1988 years, for a total of 18 year-long runs. One set of forecasts uses the original land surface parameterization, the second set uses the SiB2 land surface scheme with year specific prescribed vegetation greenness, and the third set uses the same land surface scheme with the average seasonal cycle of vegetation greenness. All three sets of forecasts use prescribed sea surface temperatures of the corresponding year. The ensemble members are initialized with the observed atmospheric conditions of January 1, 2 and 3 of the corresponding year. In the model runs using the new land surface scheme, the soil moisture is initialized as a function of the prescribed vegetation greenness for the particular set.; The effect of the treatment of land surface processes on the model climate is assessed by comparing the differences between the runs with and without the new land surface parameterization. It is found that the simulated model climate benefits from the inclusion of the more sophisticated surface treatment. The biases of the control in the near surface temperatures and humidity and the rainfall on monthly mean time scales are partially compensated for with the implementation of the new land surface scheme. Seasonal mean rainfall correlation is improved over most domains. Similarly, there is improved correlation of the difference in precipitation between the two summers to the corresponding observed difference. It is concluded that the majority of the improvement is due to the improved parameterization of the new land surface scheme rather than to the interannual variability of the vegetation greenness boundary conditions.
Keywords/Search Tags:Land surface, Parameterization, Seasonal forecast, Vegetation, Greenness
Related items