Changes in Water Resources Systems: Methodologies to Maintain Water Security and Ensure Integrated Management (Proceedings of Symposium HS3006 at IUGG2007, Perugia, July 2007).  IAHS Publ. 315, 2007, 157-171

 

Drought prediction in the Vietnamese central highlands

 

TINH DANG NGUYEN1, DAN ROSBJERG2, CINTIA UVO3 & KIM QUANG NGUYEN1

1 Faculty of Planning & Management of Water Resources Development Systems, Water Resources University, 175 Tayson, Dongda, Hanoi, Vietnam

dr@er.dtu.dk

2 Institute of Environment & Resources, Technical University of Denmark, DK-2800 Kongens Lyngby, Denmark

3 Department of Water Resources Engineering, Lund University, SE-221 00 Lund, Sweden

 

Abstract Rainfall over the Vietnamese central highlands is governed by the Asian summer monsoon. The El Niño-Southern Oscillation (ENSO) phenom­enon and related large-scale circulation anomalies, however, introduce distur­bances that may lead to drought in the central highlands. Droughts cannot be prevented, but the consequences for human livelihood and economic losses can be alleviated if better prediction tools become available. Sea surface temperature (SST) is an indicator for the ENSO phenomenon and has been used here in an effort to develop prediction models for precipitation in the Vietnamese central highlands with a lead-time of up to three months. SST in both the Indian and Pacific oceans has been related to precipitation by means of canonical correlation analysis, a linear statistical technique. The best results were obtained for rainfall at the outset and at the end of the rainy season. Nonlinear techniques in the form of artificial neural networks (ANN) were subsequently applied. Additionally, discharge in three river basins in the central highlands was predicted with SST and meteorological variables as predictors. Although local effects have a considerable influence in certain parts of the area, reasonable prediction results were obtained for both rainfall and discharge.

 

Key words  artificial neural networks; canonical correlation analysis; drought prediction; ENSO; sea surface temperature