Space-time Variability of Indonesian Rainfall at Inter-annual and Multi-decadal Time Scales

Yanto, Rajagopalan, B. & Zagona, E. (2016). “Space-time Variability of Indonesian Rainfall at Inter-annual and Multi-decadal Time Scales.” Climate Dynamics, 1-15, DOI 10.1007/s00382-016-3008-8.

Abstract
We investigated the space-time variability of wet (Nov–Apr) and dry (May–Oct) season rainfall over Indonesia, using monthly gridded rainfall data from the University of East Anglia Climatic Research Unit covering the period 1901–2012. Three complimentary techniques were employed—(1) principal component analysis to identify the dominant modes of variability, (2) wavelet spectral analysis to identify the spectral characteristics of the leading modes and their coherence with large scale climate variables and (3) Bayesian Dynamical Linear Model (BDLM) to quantify the temporal variability of the association between rainfall models and climate variables. In the dry season when the Inter Tropical Convergence Zone (ITCZ) is to the north of the equator the leading two principal components (PCs) explain close to 50% of the rainfall. In the wet season, the ITCZ moves to the south and the leading PCs explain close to 30% of the variance. El Niño Southern Oscillation (ENSO) is the driver of the leading modes of rainfall variability during both seasons. We find asymmetry in the teleconnections of ENSO to high and low rainfall years in the dry season. Furthermore, ENSO and the leading PCs of rainfall have spectral coherence in the inter-annual band (2–8 years) over the entire period of record and in the multi-decadal (8–16 years) band in post-1980 years. In addition, during the 1950–1980 period the second mode of variability in both seasons has a strong relationship with Pacific Decadal Oscillation. The association between ENSO and the leading mode of Indonesian rainfall has strengthened in recent decades, more so during the dry season. These inter-annual and multi-decadal variabilities of Indonesian rainfall modulated by Pacific climate drivers has implications for rainfall and hydrologic predictability important for water resources management.