Simple Seasonal Forecast on Monthly Precipitation for Southeast Asia Using Climate Indices
Abstract
Seasonal climate forecasts are useful for farmers and policy makers to assist in developing risk-management strategies. Dynamical predictions based on numerical simulations are widely used in practice to produce seasonal climate forecasts. However, they are expensive, especially for people in developing countries, to implement and operate. Statistical predictions based on empirical relationships between observed variables are easier to implement and operate, but there have been a few studies on statistical seasonal forecasts, although a lot of studies on concurrent relationship between climate variables and indices. The objective of the present study is to investigate the possibility of simple seasonal forecasts on monthly precipitation using climate indices. We focus on Southeast Asia, where monsoon causes remarkable seasonality of precipitation. Spline regression curves are developed for each grid in Southeast Asia to express relationships between monthly precipitations and single climate index. The CRU data, a long-term gridded climate data, was used for monthly precipitations, and 19 climate indices were used for explanatory variables. The results show that the predictability varies in seasons, regions, and lead-time. For example, precipitations in autumn in Java islands can be highly predicted with ENSO indices 2 months in advance, but precipitations in winter can not be predicted.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2019
- Bibcode:
- 2019AGUFM.B31H2411M
- Keywords:
-
- 0402 Agricultural systems;
- BIOGEOSCIENCES;
- 0466 Modeling;
- BIOGEOSCIENCES;
- 0480 Remote sensing;
- BIOGEOSCIENCES;
- 1922 Forecasting;
- INFORMATICS