Floods, cyclones, drought and climate change in Bangladesh: a reality check

Abstract

This paper seeks to correct prevailing assumptions about Bangladesh’s susceptibility to floods, tropical cyclones and drought, and the extent to which global warming has already affected the country’s climate. Analysis of 50 years of the country’s climate and hydrological data showed no evidence that rainfall amounts have changed or that floods, tropical cyclones and droughts have increased in frequency or severity. The extent to which global warming might have affected temperatures is made uncertain by the probably greater impact on temperatures at recording stations of widespread changes in land use and the heat-island effect resulting from urban expansion around the stations. The paper reviews both the diversity of environments in Bangladesh’s coastal area exposed to sea-level rise and the possible mitigation methods. Two major conclusions are drawn: that population increase and rapid urbanisation pose more serious immediate problems for development planning in Bangladesh than climate change; and that education at all levels needs to include practical field studies that could provide all students with a better understanding of the country’s diverse and locally complex environments.

Keywords:

• Bangladesh
• Climate change
• Cyclones
• Drought
• Floods

Acknowledgement

The author is grateful to Dr Robert Brinkman for constructive comments on an earlier draft of this paper.

Notes

1. The soil surveys referred to above showed that the Ganges and Brahmaputra Floodplains were not formed directly by those rivers but by smaller tributary and distributary rivers of the two major rivers. That is shown by the size and shape of the meander scars and ox-bow lakes on the floodplains. Although classified as meander floodplains, they are the meander floodplains only of the small rivers that cross them. The floodplains as a whole are the deltaic floodplains of the two major rivers.

2. The storm-surge height reported in some of the earlier cyclone occurrences included in Table 2 probably includes the maximum height of accompanying waves hitting the coast, not just the sustained height of the surge.

3. During 28 years in East Pakistan/Bangladesh, the author witnessed two megatornadoes − up to ca 1 km in diameter − which affected areas immediately east of Dhaka in 1963 and 1976, causing several thousand casualties, but he never witnessed an ordinary tornado, despite frequent travels about the country on field surveys and inspection visits during the pre-monsoon season, and he only once saw damage to a village caused by a tornado. He experienced many kal baishakhi, both in Dhaka and in the field, three with hail stones golf ball to small orange in size which killed cattle, damaged crops, broke windows and dented cars.

4. The Mymensingh meteorological station was moved from the town centre to the Bangladesh Agricultural University campus outside the town in the 1970s. The Sribardi site was moved several kilometres from a former airfield on the plains to the Bangladesh Tea Research Station on adjoining low hills during the same period. The linking of temperature data before and after these moves is invalid, therefore.

5. The author was unable to obtain information on changes in station sites. He knows from his own experience that five station sites were changed in the past 50 years, but he does not have the dates; and step changes in temperatures following breaks in the record suggest that at least another five station sites were changed.

6. Islam and Neelim [23] acknowledge that city areas can be 1−2⁰C warmer than adjoining rural areas, but they did not take this factor into account in their data analyses and interpretation; nor did they consider the possible impact on temperatures of land use changes or of site changes at some meteorological stations. Mondal, Islam and Madhu [24], in a 50-year analysis of Bangladesh’s mean annual temperatures, found a 2.4 °C per century rate of increase in mean temperature 1980–2010 that was twice that of the long-term trend (1948–2010) and projected this trend to increase in future to 4.6 °C per century, but they did not take urban warming and site shifts into account in their analysis.

7. The moribund Ganges delta comprises the area in the south-west of Bangladesh and in West Bengal where Ganges river distributaries are no longer connected with that river or a major distributary in the dry season. In this area, river salinity moves much further inland in the dry season than it does in central and eastern parts of the Ganges Tidal Floodplain where there is sufficient dry-season river flow to hold back saline intrusion to a narrow belt near the coast: see Brammer [25], figure 5.