1 Regional Hydro–Meteorological Center of Red delta river basin;;

2 Control Automation Production Institute of Technology (CAPIT);

*Corresponding author:; Tel.: +84–912509932


The research applies self-organizing maps (SOM) technique in combination with K–Means method to objectively classify weather patterns that cause summer heat wave in Viet Nam based on the dataset from 1998 to 2018. The pressure of mean sea level (PMSL) and geopotential height at 500hpa (H500) of JRA25 reanalysis data are used. The heat wave is defined to occur if the daily maximum temperature of at least 2/3 of surface synoptic stations in research area was greater than 35oC. According to above mentioned criteria, 156 summer heat waves were subjectively found at Northern region in period of 1998–2018. In central and southern regions, the summer heat waves were respectively found 204 and 69. By applying SOM and K–Means, there were 4, 3 and 2 key weather patterns that caused summer heat waves in Northern, Central and Sothern region respectively. In fact, the weather pattern caused summer heat waves at research region is usually related to activities of the western hot depression pattern and Northwest Pacific Subtropical High Pressure. The combination of 2 weather patterns or more was usually found Northern and Central region. However, the number of heat wave detected by SOM is smaller than number of heat wave was subjectively determined by forecaster (there are 109, 171 and 62 heat waves detected by SOM for the Northern, Central and Sothern region respectively). The reason for this result is that SOM method has not been able to identify heat waves caused by the combination of many weather patterns or by small and meso–scale weather patterns.


Cite this paper

Tuyet, M.T.T.; Hoa, V.V.; Danh, T.L. Application of self-organizing maps and K–Means methods to classify summer heat wave weather patterns in Viet Nam. VN J. Hydrometeorol. 2022, 11, 15-25. 



1. Ha, H.T.M.; Tan, P.V. The trend and variation of extreme temperatures in Viet Nam in the period of 1961–2007. Sci. Technol. J. Ha Noi Univ. Sc. 2009, 412–422.

2. Lanh, N.V.; Thai, D.V. Prediction of maximum temperature and heat wave in Ha Noi in May–August months under the influence of western hot depression. VN J. Hydrometeorol. 2002, 502, 14–19.

3. Lanh, N.V. The impact of the depression patterns to summer weather in Viet Nam. VN J. Hydrometeorol. 2010, 593, 43–53.

4. Lanh, N.V. 2010. Heat wave and key weather patterns that cause heat wave in Viet Nam. VN J. Hydrometeorol. 2010, 597, 1–8.

5. Hang, P.M.; Dung, T.T.; Quang, N.D. The impact of western hot depression and Northwest Pacific Subtropical High Pressure to heat wave activities over Northern Central region during 2010–2015. VN J. Hydrometeorol. 2017, 674, 44–52.

6. Lan, H.P.; Dien, N.H. Prediction of maximum temperature by using ANN for northern delta region. VN J. Hydrometeorol. 2008, 571, 20–23.

7. Hoa, V.V.; Tien, D.D.; Duc, T.A.; Hung, M.K.; Quan, D.D.; Khiem, N.V.; An, N.V. Research on classifying typical synoptic patterns causing abnormal warm spells in early winter in northern area of Viet Nam by a Self–Organizing Map. VN J. Hydrometeorol. 2019, 703, 51–59.

8. Nishiyama, K.; Endo, S.; Jinno, K.; Uvo, C.B.; Olsson, J.; Bertndtsson, R. Idenfication of typical synoptic patterns causing heavy rainfall in the rainy season in Japan by a Self–Organizing Map. Atmos. Res. 2007, 83, 185–200.

9. Liu, Y.; Weisberg, RH. Sea surface temperature patterns on the West Florida shelf using growing hierarchical Self–Organizing Maps. J. Atmos. Ocean Tech. 2005, 23, 325–338.

10. Duc, T.A. Research on classifying heavy rainfall patterns in Viet Nam using Self–Organizing Map method. Climatology and Meteorology Thesis in master degree, Ha Noi University of Science, 2014, pp. 80.

11. Tuan, V.A. etc. Research on objectively classifying heavy rainfall patterns in Viet Nam. Sci. Rep. 2015, pp. 179.

12. Seung–Yoon, B.; Kim, S.W.; Jung, M.I.; Roh, J.W.; Son, S.W. Classification of Heat Wave Events in Seoul using Self–Organizing Map. J. Clim. Change Res. 2018, 9, 209–221. Doi:10.15531/ksccr.2018.9.3.209.