Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Comparative Evaluation of Two Ensembles for Long-Range Forecasting of Monsoon Rainfall

Comparative Evaluation of Two Ensembles for Long-Range Forecasting of Monsoon Rainfall It is now well known that changes in initial conditions can give rise to substantial changes in the forecasts even at the long range. Ensemble averaging of forecasts from different initial conditions provides an efficient way of assessing and reducing uncertainties in the forecasts due to inherent uncertainties in the initial conditions. However, the procedure for generating the ensemble of forecasts has to be based on careful consideration. Although there now exist several well-tested frameworks for ensemble forecasting at the short range, the procedure for and impact of ensemble forecasting on long-range forecasting of monsoons remain relatively less explored. In particular, the procedure for the choice of the ensemble for long-range forecasting of monsoons needs special considerations. The Indian summer monsoon is characterized by a number of intraseasonal oscillations (ISOs) whose phases and amplitudes can significantly affect the monsoon forecast and that can be adequately sampled only using initial states spread over time scales comparable to the characteristic time scales of these ISOs. It is shown that use of initial states spread over a longer period (such as 1 April–1 May) results in a better ensemble average for long-range forecasting of Indian summer monsoon than that from an ensemble of closely packed states with a shorter lead. An optimized configuration for long-range forecasting of monsoons using a variable resolution general circulation model is adopted. The climatological monthly mean SST field is used to assess realizable skill, as use of observed SST provides only potential skill. Then five-member wide-lead (1 April–1 May) ensemble average forecasts are compared with five-member compact-lead (27 April–1 May) ensemble average forecasts for 24 (1980–2003) hindcasts; it is shown that the skill of the wide-lead ensemble average is superior to that of the compact-lead ensemble at different spatial and temporal scales in spite of the longer lead of the former. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Monthly Weather Review American Meteorological Society

Comparative Evaluation of Two Ensembles for Long-Range Forecasting of Monsoon Rainfall

Monthly Weather Review , Volume 137 (9) – Aug 26, 2008

Loading next page...
 
/lp/american-meteorological-society/comparative-evaluation-of-two-ensembles-for-long-range-forecasting-of-Q1tBSgnUjo

References (73)

Publisher
American Meteorological Society
Copyright
Copyright © 2008 American Meteorological Society
ISSN
1520-0493
DOI
10.1175/2009MWR2767.1
Publisher site
See Article on Publisher Site

Abstract

It is now well known that changes in initial conditions can give rise to substantial changes in the forecasts even at the long range. Ensemble averaging of forecasts from different initial conditions provides an efficient way of assessing and reducing uncertainties in the forecasts due to inherent uncertainties in the initial conditions. However, the procedure for generating the ensemble of forecasts has to be based on careful consideration. Although there now exist several well-tested frameworks for ensemble forecasting at the short range, the procedure for and impact of ensemble forecasting on long-range forecasting of monsoons remain relatively less explored. In particular, the procedure for the choice of the ensemble for long-range forecasting of monsoons needs special considerations. The Indian summer monsoon is characterized by a number of intraseasonal oscillations (ISOs) whose phases and amplitudes can significantly affect the monsoon forecast and that can be adequately sampled only using initial states spread over time scales comparable to the characteristic time scales of these ISOs. It is shown that use of initial states spread over a longer period (such as 1 April–1 May) results in a better ensemble average for long-range forecasting of Indian summer monsoon than that from an ensemble of closely packed states with a shorter lead. An optimized configuration for long-range forecasting of monsoons using a variable resolution general circulation model is adopted. The climatological monthly mean SST field is used to assess realizable skill, as use of observed SST provides only potential skill. Then five-member wide-lead (1 April–1 May) ensemble average forecasts are compared with five-member compact-lead (27 April–1 May) ensemble average forecasts for 24 (1980–2003) hindcasts; it is shown that the skill of the wide-lead ensemble average is superior to that of the compact-lead ensemble at different spatial and temporal scales in spite of the longer lead of the former.

Journal

Monthly Weather ReviewAmerican Meteorological Society

Published: Aug 26, 2008

There are no references for this article.