We found a match
Your institution may have rights to this item. Sign in to continue.
- Title
Mechanism for high "potential skill" of Indian summer monsoon rainfall prediction up to two years in advance.
- Authors
Sharma, Devabrat; Das, Santu; Saha, Subodh K.; Goswami, B. N.
- Abstract
Skillful prediction of the Indian summer monsoon rainfall (ISMR) at leads of 12–24 months are valuable for farmers and policymakers for water resource management and food security planning. While the ISMR is known to be highly predictable at short leads, estimates of long‐lead "potential skill" is lacking. Here, a new predictor discovery method taking into account simultaneous contributions from all potential drivers unravels a predictor based on the depth of the 20° isotherm (D20) to be best suited for estimating the "potential skill" significantly better than a sea surface temperature (SST)‐based predictor. We find high "potential skills" at 18–24‐months leads instead of at short leads, with the highest skill for 18‐months‐lead forecasts (r = 0.87). Investigation of the intriguing finding reveals that the initial errors for ISMR prediction as well as their growth are controlled by a global manifestation of the El Niño–Southern Oscillation (ENSO) oscillator. Phase‐locking of initial errors of D20 and their growth rates with the annual cycle ascertain that forecast errors are minimum and "potential skill" is maximum at leads of 18–24 months. While non‐linearity of small‐scale D20 anomalies may make realization of the potential predictability challenging, with improvements of coupled models and use of deep‐learning artificial‐intelligence techniques, our findings provide optimism for skillful long‐lead forecasts of ISMR in coming years.
- Subjects
RAINFALL; WATER management; EL Nino; OCEAN temperature; MONSOONS; FORECASTING
- Publication
Quarterly Journal of the Royal Meteorological Society, 2022, Vol 148, Issue 749, p3591
- ISSN
0035-9009
- Publication type
Article
- DOI
10.1002/qj.4375