Premium
Skill of monthly rainfall forecasts over India using multi‐model ensemble schemes
Author(s) -
Kar Sarat C.,
Acharya Nachiketa,
Mohanty U. C.,
Kulkarni Makarand A.
Publication year - 2011
Publication title -
international journal of climatology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.58
H-Index - 166
eISSN - 1097-0088
pISSN - 0899-8418
DOI - 10.1002/joc.2334
Subject(s) - climatology , probabilistic logic , forecast skill , environmental science , monsoon , meteorology , statistical model , ensemble average , scale (ratio) , distribution (mathematics) , ensemble forecasting , statistics , mathematics , geography , geology , cartography , mathematical analysis
Rainfall in the month of July in India is decided by large‐scale monsoon pattern in seasonal to interannual timescales as well as intraseasonal oscillations. India receives maximum rainfall during July and August. Global dynamic models (either atmosphere only or coupled models) have varying skills in predicting the monthly rainfall over India during July. Multi‐model ensemble (MME) methods have been utilized to evaluate the skills of five global model predictions for 1982–2004. The objective has been to develop a prediction system to be used in real time to derive the mean of the forecast distribution of monthly rainfall. It has been found that the weighted multi‐model ensemble (MME) schemes have higher skill in predicting July rainfall compared to individual models. Through the MME methods, skill of rainfall predictions improved significantly over eastern parts of India. However, there is a region over India where none of the models or the MME scheme has any useful skill. Similarly, there are few typical years in which the mean distribution of July rainfall cannot be predicted with higher skill using the available statistical post‐processing methods. A simple MME probabilistic scheme has been utilized to show that skill of probabilistic predictions improved when the representation of mean of forecast distribution has better skill. Copyright © 2011 Royal Meteorological Society