Goal

· Development of multi-model based regional seasonal prediction techniques

· Analysis of subseasonal variability and predictability

· Development of multi-model based subseasonal prediction techniques

Basic Functions

· Development of multi-model based climate prediction techniques

· Diagnostics and verification of subseasonal-to-seasonal climate prediction

· Assessment of subseasonal-to-seasonal climate predictability

· Development of high-resolution, gridded climate information on a regional scale

Main Research

Seasonal prediction generally aims to predict a monthly average of the climate for the following three to six months. To generate reliable seasonal predictions, it is essential to have not only a great number and diversity of climate model predictions, but also advanced ensemble techniques, through which initial and model uncertainties can be addressed. The Prediction Research Department (PRD) analyzes the major modes of climate variability, assesses the corresponding prediction skill of global climate models, and eventually develops multi-model ensemble techniques to provide improved seasonal prediction. PRD also investigates multi-model based regional seasonal prediction techniques for the Asia-Pacific region (including South Korea).

Extreme climate events such as heat waves, cold surge, heavy rain, and droughts, have profound socio-economic impacts. These events are more often associated with atmospheric phenomena at the subseasonal timescale (a time span of about 15-60 days). With accumulated expertise in multi-model ensemble seasonal prediction, PRD is working on assessing subseasonal predictability, diagnostics, and verification of climate model subseasonal prediction, and developing multi-model based subseasonal prediction techniques. PRD also explores the application of cutting-edge technologies (for example, deep learning techniques) to climate model subseasonal prediction to improve prediction skill.